(navigation image)
Home American Libraries | Canadian Libraries | Universal Library | Community Texts | Project Gutenberg | Children's Library | Biodiversity Heritage Library | Additional Collections
Search: Advanced Search
Anonymous User (login or join us)
Upload
See other formats

Full text of "The 2nd NASA Aerospace Pyrotechnic Systems Workshop"

Ci7< '?>. 



NASA Conference Publication 3258 



Second NASA Aerospace 
Pyrotechnic Systems 

Workshop 



Compiled by 

William W. St. Cyr 

John C. Stennis Space Center 

Stennis Space Center, Mississippi 



Proceedings of a workshop sponsored by the 

Pyrotechnically Actuated Systems Program 

Office of Safety and Mission Quality 

National Aeronautics and Space Administration 

Washington, D.C., and held at 

Sandia National Laboratories 

Albuquerque, New Mexico 

February 8-9, 1994 




National Aeronautics and 
Space Administration 

Office of Management 

Scientific and Technical 
Information Prgram 

1994 



Workshop Management 

NASA Pyrotechnically Actuated Systems Program 

Norman R. Schulze, NASA Headquarters - Program Manager 

Host Representative 

Jere Harlan - Sandia National Laboratories 

Workshop Coordinator 
William W. St. Cyr - NASA, John C. Stennis Space Center 



Workshop Program Committee 



Norman R. Schulze 

National Aeronautics and Space Administration 

Code QW 

Washington, DC 20546 



James Gageby 

The Aerospace Corporation 

P. 0. Box 92957 

Los Angeles, CA 90009 



Laurence J. Bement 

National Aeronautics and Space Administration 

Langley Research Center 

Code 433 

Hampton, VA 23665 



Anthony Agajanian 

Jet Propulsion Laboratory 

Code 158-224 

4800 Oak Grove Drive 

Pasadena, CA 91109 



William W. St. Cyr 

National Aeronautics and Space Administration 

John C. Stennis Space Center 

Code KA60 - Building 1 100 

Stennis Space Center, MS 39529 



Table of Contents 



WELCOMING ADDRESS 1 " 9 

Dr. John Stichman, Director, Surety Components and Instrumentation Center, 
Sandia National Laboratories 

NASA Pyrotechnically Actuated Systems Program 1 3 " ^ I 

Norman R. Schulze, NASA, Washington DC 



SESSION 1 - Laser Initiation and Laser Systems 

Laser Initiated Ordnance Activities in NASA 29 ^ d 

Norman R. Schulze, NASA, Washington DC 

Laser-Ignited Explosive and Pyrotechnic Components 49 - Q 

Al Munger, Tom Beckman, Dan Kramer and Ed Spangler, 
EG&G Mound Applied Technology 

A Low Cost Ignitor Utilizing an SCB and Titanium Sub-Hydride Potassium Perchlorate 
Pyrotechnic 61—^5 

Robert W.Bickes, Jr. and M. C. Grubelich, Sandia National Laboratories 

J. K. Hartman and C. B. McCampbell, SCB Technologies, Inc. 

J. K Churchill, Quantic-Holex 

Optical Ordnance System for Use in Explosive Ordnance Disposal Activities 65 - 4~ 

John A. Merson, F. J. Sa/as and F. M. Helsel, Sandia National Laboratories 



r- 



Laser Diode Ignition 71"*^ 

William J. Kass, Larry A. Andrews, Craig M. Boney, Weng W. Chow, 
James W. Clements, Chris Co/burn, Scott C. Holswade, John A. Merson, 
Fredrick J. Salas, and Randy J. Williams, Sandia National Laboratories 
Lane Hinkle, Martin Marietta Specialty Components 

Standardized Laser Initiated Ordnance System 79 — £ 

James V. Gageby, The Aerospace Corporation 

Miniature Laser Ignited Bellows Motor 83 ~ f 

Steven L. Renfro, The Ensign-Bickford Co. 

Performance Characteristics of a Laser-Initiated NASA Standard Initiator 93 - % 

John A. Graham, The Ensign-Bickford Co. 

Four Channel Laser Firing Unit (LFU) Using Laser Diodes 101 \ 

David Rosner and Edwin Spomer 

Pacific Scientific, Energy Dynamics Division 

LIO Validation on Pegasus (Oral Presentation Only) 115 — Q 

Arthur D. Rhea, The Ensign-Bickford Co. 



in 



SESSION 2 - Electric Initiation 

EBWs and EFIs - The Other Electric Detonators 117 - I 

Ron Varosh, Reynolds Industries Systems, Inc. 

Low Cost, Combined RF and Electrostatic Protection for Electroexplosive Devices ... 125 - '/ 
Robert L. Dow, Attenuation Technology, Inc. 

Unique Passive Diagnostic for Slapper Detonators 131 ™ * 1 

William P. Brigham and John J. Schwartz, Sandia National Laboratories 

Applying Analog Integrated Circuits for HERO Protection 143 - \\ 

Kenneth E. Willis, Quantic Industries, Inc. 
Thomas J. Blachowski, NSWC, Indian Head MD 

Cable Discharge System for Fundamental Detonator Studies 149 h 

Steven G. Barnhart, Gregg R. Peevy and William P. Brigham, 
Sandia National Laboratories 

Improved Test Method for Hot Bridgewire AII-fire/No-fire Data 165 ^ (S 

Gerald L. O'Barr, Retired (formerly with General Dynamics) 

SESSION 3 - Mechanisms & Explosively Actuated Devices 

Development and Demonstration of an NSI Derived Gas Generating Cartridge (NSGG) 177 ~lQ 
Laurence J. Bement, NASA, Langley Research Center 
Harold Karp, Hi Shear Technology Corp. 
Michael C. Magenot, Universal Propulsion Co. 
Morry L. Schimmel, Schimmel Company 

Development of the Toggle Deployment Mechanism 191 -it 

Christopher W. Brown, NASA, Johnson Space Center 

The Ordnance Transfer Interrupter, A New Type of S&A Device 213 ^ 

John T. Greensfade, Pacific Scientific, Energy Dynamics Division 

A Very Low Shock Alternative to Conventional, Pyrotechnically Operated Release 

Devices 223 ~~!^ 

Steven P. Robinson, Boeing Defense & Space Group 

Investigation of Failure to Separate an Inconel 718 Frangible Nut 233 ~ ^O 

William C. Hoffman III and Carl W. Hohmann, NASA, Johnson Space Center 



SESSION 4 - Analytical Methods & Studies 

Bolt Cutter Functional Evaluation 243 

S. Goldstein, S. W. Frost J- B. Gageby, T. E. Wong, and R. B. Pan, 
The Aerospace Corporation 

Choked Flow Effects in the NSI Driven Pin-Puller 269 

Joseph M. Powers and Keith A. Gonthier, University of Notre Dame 



IV 



-2£ 



Finite Element Analysis of the 2.5 Inch Frangible Nut for the Space Shuttle 285 ^Z ^ 

Darin McKinnis, NASA, Johnson Space Center 

Analysis of a Simplified Frangible Joint System 297 " £ 4- 

Steven L Renfro and James E. Fritz, The Ensign-Bickford Co. 
Steven L Olson, Orbital Sciences Corp. 

Portable, Solid State, Fiber Optic Coupled VISAR for High Speed Motion and Shock 

Diagnostics 309 

Kevin J. Fleming and O. B. Crump, Jr., Sandia National Laboratories 

Development and Qualification of a Laser-Ignited, All-Secondary (DDT) Detonator ... 317 ~ 2 .Q> 
Steve Tipton, Air Logistics Center (AFMC) 
Thomas J. Blachowski and Darrin Krivitsky, NSWC, Indian Head MD 



SESSION 5 - Miscellaneous « 

Pyrotechnically Actuated Systems Database and Catalog 331 

Paul Steffes, Analex Corporation 

Fire As I've Seen It 337 

Dick Stresau, Stresau Laboratory 



SESSION 6 - Panel Discussion/Open Forum 

Moderator: William W. St. Cyr, NASA, John C. Stennis Space Center 

Panel Members: 

Norm Schulze, NASA Headquarters - Code QW 
Larry Bement, NASA Langley Research Center 
Tom Seeholzer, NASA Lewis Research Center 
Jere Harlan, Sandia National Laboratories 
Jim Gageby, The Aerospace Corporation 
Ken VonDerAhe, Pacific Scientific Corporation 

Discussion topics: 

Topics submitted from the audience. 

What's the future of pyrotechnics? 

What are the predominant issues pertaining to pyrotechnics? 

Pyrotechnic failures - lessons learned. 

Pyrotechnic coordination - is it needed? 

Are standards and standard specifications needed? 

Note: The panel discussion and open forum commenets were not recorded and are not 
included in this conference publication. 



APPENDIX A - List of Participants A-1 



omit 



1994 NASA AEROSPACE 
PYROTECHNIC SYSTEMS WORKSHOP 

February 8-9, 1994 

Sandia National Laboratories 

Albuquerque, NM 



Welcoming Address - John Stichman, Director 

Surety components and Instrumentation Center 

Sandia National Laboratories 



■ 1 - 



'paceJA..,-.- 



Welcome 



■ 




Primary purpose of this conference is to promote communications 
between government agencies and private industry. 




Conference is Organized in Four Sessions 



■ 

CO 




Optical Ordnance 

- New, exciting and potentially significant safety improvements. 

Electrical Initiation 

- Backbone of pyrotechnics. 

- Continued safety improvements (ESD & EMR). 

Mechanisms & Explosive Activated Devices 

- Details of device designs are important for specific technology 
transfer. 

Analytic Methods & Studies 

- Computer analysis & improved test methods lead to a deeper 
understanding & improved performance characterization 




Five Decades of National Service 



4* 




-x;>: : :%^ ....;. ■$■; '/'. ^.| : '" :: ".:•:. ..'.-. 



National Laboratories have had opportunity & privilege to support 
U.S. nuclear weapons programs where we must provide long life & 
high reliability components. Performance is characterized by 
studies of fundamental mechanisms of explosive & pyrotechnic 
phenomena. 




Sandia has two major laboratory locations 



Cl 




New Mexico 



California 



The Defense Programs Sector is responsible for a 
significant fraction of the Laboratories' activities 



o> 




Other 



Intelligence 

Nuclear 
safeguards 
and security 



Research and 
development 




Nuclear 
testing 



Technology 
transfer 
initiative 



Verification and 
control 
technology Stockpile 

support 



Inertial 

confinement 

fusion 





m 



Our programmatic efforts address 
changing national needs 



DOLLARS IN MILLIONS 
1400 







70 72 74 76 78 80 82 84 86 88 90 92 

Fiscal Years 



Engineering and physical sciences are the 
backbone of our multidiscipline laboratories 




61 Support 

D Other Professionals 

□ Technicians 

M Electrical Engineering 

H Mechanical Engineering 



H Computer Science 
H Other Engineering 
M Physics 
H Chemistry 
IS Other Science 



Modern process development and 
prototype fabrication facilities 



■ 



iProcess 

[Development 

Laboratory 



Manufacturing 
Technologies Facility 




Integrated 
Manufacturing 
Technologies 
Laboratory 






Improved Industrial 
Competitiveness 



o 

I 




mmmmmmm 

: .- ----- - . ■ -: ---- -■ - 



Technology transfer in the past has been controlled 
for security reasons, however, today it is becoming a 
mission of the national laboratories to help other 
government agencies and private industry in order to 
improve our economic security. 




Sandia undertakes new programs that enhance national 
security and industrial competitiveness 




National security 

• Military 

• Energy 

• Economic 

• Environmental 



1 


a * 

r Mi; 






i 



Teaming with industry 




"All federal R&D agencies will be encouraged to 
act as partners with industry wherever possible. 
All laboratories managed by the Department of 
Defense that can make a productive contribution 
to the civilian economy will be reviewed with the 
aim of devoting at least 10-20 percent of their 
budgets to R&D partnerships with industry. In this 
L way, federal investments can be managed to 
? benefit both government's needs and the needs 
of U.S. Business" 

President Bill Clinton 

"Technology for America's Economic Growth" 

February 22, 1993 




si-si 



f' 1 



V 




Update: NASA 

Pyrotechnically Actuated 

Systems Program 

Norman R. Schulze, NASA Headquarters 

Second NASA Aerospace Pyrotechnic Systems Workshop 

Sandia National Laboratory, Albuquerque, NM 

February 8, 1994 



13 



=0$MA* 



February 8, 1994 



=mMA= 



Agenda 



I. Program Origin 

II. Program Description 

III. Summary 



I. Program Origin 



14 



Introduction - Pyrotechnic Systems 

=mM A 



• Routinely perform wide variety of mechanical functions: 

- Staging 

- Jettison 
Control flow 
Escape 

- Severance 

• Mission Critical 

• Are required to have near perfect reliability 

• But failures continue, some repeatedly 



February 8, 1994 



Definition 



By example, pyrotechnic devices and systems include: 

- Ignition devices 

- Explosive charges and trains ! 

- Functional component assemblies, e.g., pin pullers, cutters, explosive j 
valves, escape systems ! 

- Systems, i.e., component assembly, ignition circuitry, plus the interactions 
with the environment such as structure, radio waves, etc. 



February 8, 1994 - 1 5 - 



—osmA' 



Summary of Survey 



23 year span covered 
Failure categories 

- Initiation 

- Mechanisms 

- Spacecraft separation systems and linear explosives 
Firing circuits 

Reviewed by Steering Committee 
Report prepared 

- Bement, L. J., "Pyrotechnic System Failures: Causes and Prevention," 
NASA TM 100633, Langley Research Center, Hampton, VA, June 1988 



February 8, 1994 



—0$MA* 



Assessment of Survey Results 



Deficient Areas 


Recommended Tasks 


• Design Approaches 

- generic specification 

- standard devices 


• Design Approaches 

- prepare NASA specification handbook 

- select/verify existing hardware types 


• Pyrotechnic Technology 

- research/development technology base 

- recognized engineering discipline 

- training/education 

- test methodology/capabilities 

- new standard hardware 


• Pyrotechnic Technology 

- endorse and fund plan's technology tasks 

- fund training and academic efforts 

- R&D for new measurement techniques 

- develop new h/w for standard applications 


• Communications 

- technology exchange 

- data bank & lessons learned 

- intercenter program support 


• Communications 

- continue Steering Committee meetings 

- initiate symposia 

- establish pyro reporting requirements for NASA PRACA 

- perform as a Steering Committee function 


• Resources 

- funds 

- research/development staff and facilities 


• Resources 

- implement pyrotechnic program plan 



February 8, 1994 



■ 16 



—0SMA* 



II. Program 



• PAS Program Goals 

• Program Flow 

• PAS Program Organization 



1.0 Program Requirements and 
Assessments Element 



• Implement projects necessary to address management aspects of the 
Program's objectives 

• Emphasize documentation and communications 

• Prepare policy and planning documents to ensure products used 

• Analyze NASA's future program requirements and current problems 

• Provide computerized data base 

• Produce documentation related to reviews, proceedings, analyses, etc. 



February 8, 1994 



- 17 



1.1 Future Pyrotechnic Requirements 



Project Mgr: N. Schulze, Headquarters 

• Determine new pyrotechnic technology requirements 

• Define efforts to: 

- Improve PAS quality 

- Meet more demanding environments 

- Extend service requirements 

• Evaluate new diagnostic techniques 

• Provide functional understanding using computational modeling 
capabilities - enhance specifications 

• Product: 

- Report on analysis of future requirements 
STATUS: 

On hold pending program review 



February 8, 1994 10 



1 .3 PAS Technical Specification 



Project Mgr: B. Wittschen, Johnson Space Center 

• Develop common procurement specifications 

• Provide consistent technical reference for common technologies 

• Use shared experience 

• Make applicable to design, development, demonstration, environmental 
qualification, lot acceptance testing, and documentation 

• Assure critical concerns addressed using expertise of pyro community 

• Provide common in-process quality assurance measures 

• Product: 

- NASA Handbook (NHB) 
STATUS: 

On hold pending action by Pyrotechnic Steering Committee to complete 
review of the document 



February 8, 1994 - 1 ft - 1 1 



1 .4 PAS Data Base 



Project Mgr: T. Seeholzer, Lewis Research Center 

• Include past and current programs in terms of a hardware database 
incorporating system requirements, designs developed, performance 
achieved, specifications, lessons learned, and qualification status 

• Present sufficient detail to provide guidance for users 
Pyrotechnic Catalogue: 

• Describe PAS devices used on prior programs 

• Make available single data source to provide information on applications 
of pyrotechnic devices including: 

- their requirements 

- physical envelopes 

- weights 

- functional performance 

- lessons learned 

- environmental qualification 

- flight history 



February 8. 1994 



12 



1.4 PAS Data Base (continued) 



• Provide available information on pyrotechnic flight failures 

• Coordinate with industry 

• Product: 

- Catalogue to be made available upon request 
STATUS: 

• Project is underway, content selected, data being complied, first draft 
submitted to Committee for review, comments being incorporated 

• Workshop paper to provide details 

• Project completion expected in 1995 



February 8, 1994 ■ 1Q ■ 13 



1 .7 NASA PAS Manual 

=mMA 



Project Mgr: L. Bement, Langley Research Center 

• Develop detailed "how-to" document to provide guidance on all aspects 
of design, development, demonstration, qualification (environmental), 
common test methods, margin demonstrations, etc. of pyrotechnically 
actuated devices and systems 

• Scope: Applies to pyro life cycle from creation of PAS/component 
design to final disposition of device 

• Product: 

- NASA Handbook (NHB) for reference 
STATUS: 

• Project is underway, content selected, text/data being complied 

• Project completion expected in approximately one year 



February 8,1994 14 

1 .8 Pyrotechnically Actuated Systems 

Workshop 



Project Mgr: W. St. Cyr, Stennis Space Center 

• Create opportunity for technology exchanges at national level 

• Perform planning for review by the Steering Committee 

• Presentations by government and industry personnel on latest 
developments 

• Informal to facilitate communications 

• Product: 

- Workshop organization, preparations, implementation, and preparation of 
proceedings in a timely manner 
STATUS: 

• First Workshop held on June 9-10, 1992 

• Workshop proceedings published and distributed, NASA CP-3169 



February 8,1994 - 20 - 15 



2.0 Design Methodology Program Element 



• Applied technology focus 

• Hardware developed 

• Emphasize design standards and analytical techniques 

• Decrease chance of failure of new hardware design approaches or of 
proven hardware in new operational regimes 

• All aspects of pyrotechnic component and systems applications 
covered 

• Provide guidelines, handbooks, and specifications for design and 
development of pyrotechnic components and systems 



February 8, 1994 15 



2. 1 NASA Standard Gas Generator (NSGG) 

=Q3MA : : = = 



Project Mgr: L. Bement, Langley Research Center 

• Develop where the use of gas output is needed to perform a function 
rather than serving as ignitor: 

- Separation nuts, valves, cutters, switches, pin pullers, thrusters, mortars, 
bolts, etc. 

• Common NASA GG 

- Based on NSI (NASA Standard Initiator) to provide pedigree 

- Important for safety 

- Saves $, NSI 

- Wide variety of cartridges - lack "pedigree" inherent with a "Standard" 



February 8, 1994 17 

- 21 - 



2.1 NSGG (continued) 

=mM A — = 



• Develop sizes to meet wide range of performance requirements 

• Products: 

- Qualified NSGG 

- Design specification (NHB) 

- Test reports 
STATUS: 

• Project has been successfully completed 

• Two sources 

• Workshop paper to provide details 



February 8, 1994 18 



2.2.1 NASA Standard Linear Separation 

System (NSLSS) 



Project Mgr: Joe B. Davis. Marshall Space Flight Center 

• Develop standard linear separation system 

- Improved, more reliable, high performance hardware 

- Lower cost 

• Characterize functional performance, effects of system variables, 
including scaling 

• Specify process controls to assure consistency and reliability 

• Qual test for flight 

• Establish operational functional margin 

• Solicit design approaches from industry 

- Prepare NASA-wide technical specification 
STATUS: 

• Project has been terminated due to lack of funds 



February 8, 1994 _ «0 _ 19 



2.5 Advanced Pyrotechnically Actuated 

Systems (PAS) 



=mMA= 



• Define and pursue advanced design concepts to bring NASA programs 
up to the state-of-the-art in pyrotechnic technology 

• Maintain currency 
STATUS: 

• Project has been terminated due to lack of funds 



February 8, 1994 20 



3.0 Test Techniques Program Element 

=QSMA 



Address all aspects of testing: manufacturing, lot 

acceptance, qualification, margin validation, accelerated life, 

ground checkout, and in-flight checkout 

Provide better characterization of component and system 

performance 



February 8, 1994 _ aa ^ 21 



3.1 NSGG Performance 

'€>SM A 



Project Mgr: L. Bement, Langley Research Center 

• Test to demonstrate NSGG for flight 

• Develop test procedures 

• Quantify performance 

• Qualify NSGG 

• Prepare design and test specifications 

• Products: 

- Design and test specification 

- NSGG qualification test report 
STATUS: 

• Project has been successfully completed 

• Functional performance and qualification completed 

• Workshop paper to provide details 



February 8, 1994 22 



3.2 Standard System Designs 



• Provide improved, more reliable, high performance standard 
hardware designs 

• Establish functional performance, effects of system 
variables, scaling 

• Prioritized selection of candidate hardware to become 
"standards" 

• Products: 

System designs flight qualified and reports 

Process controls specified in a technical specification 



February 8, 1994 _ 24 



3.2.1 NSLSS Performance 



Project Mgr: L Bement, Langley Research Center/J. Davis, MSFC 

• Demonstrate functional performance of the NSLSS developed in Project 
2.2.1. 

• Develop test procedures for the NSLSS that confirm its intended 
operation 

• Quantify performance output and update design specifications 

• Products: 

- System design(s) flight qualified 

- Process controls specified in a technical specification 

- Comprehensive final report 
STATUS: 

• Project has been terminated due to lack of funds 



February 8, 1994 24 



3.6 Service Life Aging Evaluations 



Project Mgr: L Bement, Langley Research Center 

• Evaluate effects of aging on pyrotechnic devices and degradation from 
storage in the intended operational environments 

• Determine relationships between storage environments and device 
shelf life 

• Evaluate accelerated life test approaches I 

• Find performance characteristics that can be measured during I 
qualification to ensure that function and margins are not impaired by 

long periods of storage 

• Product: 

- Guidelines for estimating service life 



February 8, 1994 oc 25 



=€)SMA= 



3.6.1 Service Life Evaluations - 

Shuttle Flight Hardware 



Project Mgr: L. Bement, Langley Research Center 

• Determine effects of aging on Shuttle flight pyrotechnic devices 

- Ensure that function and margins not impaired by long periods of storage 

- 42 units tested 

• Compare actual space flight hardware with older hardware stored on 
the ground under controlled conditions 

• Test phase recently completed 

- Results look good. Five year extension. 
STATUS: 

• Project has been successfully completed 

• Service life extended 



February 8,1994 26 



4.0 Process Technology Program Element 



Put science into design and analysis of pyros 

Develop approaches for analytically characterizing device performance 

sensitivities to manufacturing tolerances and "faults," or deviations, in 

component ingredients 

Perform tests that verify analysis 

Address problems caused by inadequately controlled specifications or 

introduction of unanticipated substances into manufacturing process 

Establish proper degree of controls for assuring product quality and 

reliability 

Emphasize process understanding and controls to assure that specified 

hardware performance is realized during manufacturing processes 

Support product inspection criteria and acceptance testing criteria 



February 8, 1994 _ 2g . 27 



4.2 NSI Model Development 



Project Mgr: R. Stubbs, Lewis Research Center 

• Provide better understanding of NSI's sensitivities to the effects of 
process variables on performance 

• Develop model 

- Contract with Dr. J. Powers and Dr. K. Gonthier, U. of Notre Dame 

• Verify by testing 

• Present necessary technical details to control device's function 
providing consistently high reliability level of performance 

• Products: 

- Validated model 

- Report describing model in specification format 
STATUS: 

• Project has been successfully completed 

• Feasibility of modeling demonstrated 

• Workshop paper to provide details 

• Work given international recognition: International Pyrotechnics 
Society Award, to be presented February 20-25, 1994 at Christchurch 
New Zealand 

February 8. 1994 28 



Summary 

—&SMA 



• Program presented in the 1992 has been substantially phased down 

• Funded projects in work/completed as planned within cost and 
schedule constraints: 

- Data Base (in work) 

- Pyro manual (in work) 

- Workshop (no funds) 

- NSI Derived Gas Generating Cartridge 

- Shuttle Pyrotechnics Service Life Extension 

- Laser ordnance demonstration (Pegasus) (in work) 

- Laser ordnance demonstration (Shuttle Cargo Bay) (in work) 

- Modeling NSI 

- Modeling Linear Separation System (in work) 



February 8, 1994 *— 29 



Summary (continued) 

—dDSMA 



Programs eliminated: 

- Linear Separation System 

- Improved safe and arm system 

- Advanced standard hardware 

- Standard components and detonator 

- Training 

Goal was to reduce risks on future programs through better engineering 

understandings of pyrotechnic deveices 

Pyrotechnic problems persist - one of the most likely causes for the 

failure of the Mars Observer 

New program initiatives may be forthcoming as a result of that failure 

Plan to be given senior management attention 

Only advocacy at NASA Headquarters for pyrotechnics resides in Code Q 



February 8,1994 30 



- 28 







Laser Initiated Ordnance (LIO) 
Activities in NASA 



Norman R. Schulze, NASA Headquarters 

Second NASA Aerospace Systems Workshop 

Sandia National Laboratory, Albuquerque, NM 

February 8, 1994 



29 - 



Second NASA Aerospace Pyrotechnic Workshop 
Laser Initiated Ordnance Activities in NASA 



LASER INITIATED ORDNANCE BENEFITS 

[GOALS FOR ANY PROGRAM] 

• GREATER RELIABILITY 

• ENHANCED SAFETY 

• LIGHTER WEIGHT 

• LESS COSTLY PRODUCTS 

• IMPROVEMENTS IN DESIGN LEADING TO HIGHER 

OPERATIONAL EFFICIENCY 



January 30, 1994 



Second NASA Aerospace Pyrotechnic Workshop 
Laser Initiated Ordnance Activities in NASA 



APPLICATIONS 

• INITIATION OF SEQUENCING FUNCTIONS 

• FLIGHT TERMINATION 

• PROGRAM APPLICATIONS 

- new launch vehicles 

- selected use on existing fleet designs 

- spacecraft 

• LASERS HAVE LONG DEVELOPMENTAL HISTORY BUT LACK 

OPERATIONAL PEDIGREE 

— 15+ years 

- small ICBM rod lasers, first laser ordnance flight test 



3 - 30 - 



January 30. 1994 



Second NASA Aerospace Pyrotechnic Workshop 
Laser Initiated Ordnance Activities in NASA 



ADVANTAGES OF LASER ORDNANCE 

• PHYSICS OF PHOTON NOT SUSCEPTIBLE TO HAZARDS OF ELECTRON: 

ELECTROSTATICS, EMI, RF 

• LASER DIODES HAVE THE POTENTIAL FOR DESIGN OF ALL SOLID STATE 

SYSTEM 

• POTENTIAL FOR BUILT-IN-TEST (BIT) 

• PERMITS LESS SENSITIVE INITIATION ORDNANCE 

• ELIMINATES POSSIBLE HAZARD TO ELECTRONIC EQUIPMENT FROM 

FIRING OF HOT BRIDGEWIRE CARTRIDGE 

- Mars Observer failure option 

- Magellan 

• BOTTOM LINE: THE ABOVE FEATURES, WE SAY, FOR LASER DIODES 

EQUATE TO IMPROVEMENTS IN SAFETY, RELIABILITY, OPERATIONS, 
COST, POWER, MASS 

• CONCLUSION: ADDRESS LASER DIODE ORDNANCE 
DEVELOPMENT FOR OPERATIONAL FEASIBILITY 



January 30, 1994 



Second NASA Aerospace Pyrotechnic Workshop 
Laser Initiated Ordnance Activities in NASA 



DISADVANTAGES OF LASER DIODE INITIATED 

ORDNANCE 

• TECHNICAL 

- Low voltage to activate laser 

- concern over electronics setting off laser accidentally 

- BIT not proven 

- development of requirements necessary 

• MANAGERIAL: 

- Hardware not proven with operational experience 

- application not mandatory for program success 

- new programs wait for others to "break the ice" to reduce risks with cost, performance, 
schedule 

- Incomplete understanding of requirements 



5 - 31 - 



January 30, 1994 



Second NASA Aerospace Pyrotechnic Workshop 
Laser Initiated Ordnance Activities in NASA 



IN THE BEGINNING 



PAS PROGRAM PLAN 
LIO PROGRAMS 

2.4 NASA STANDARD LASER DIODE SAFE AND ARM 
2.5.1 NASA STANDARD LASER DETONATOR 
3.4 LASER DIODE SAFE/ARM PERFORMANCE 



January 30, 1994 



Second NASA Aerospace Pyrotechnic Workshop 
Laser Initiated Ordnance Activities in NASA 



PAS PROGRAM PLAN FOR LIO 

2.4 NASA STANDARD LASER DIODE SAFE AND ARM 

Project Mgr: B. Wittschen, Johnson Space Center 

• Develop, qualify, and demonstrate in flight a standardizable solid state laser safe and arm system 

- Flight demonstration - TBD 

- Joint HQS. activity with JSC 

• Determine criteria for what constitutes an acceptable S&A 

- Closely involve range safety in the design and testing 

- Place operational considerations up front in the design 

• Enhance safety and reduce risk 

- Enhance functional reliability 

- Simplify design 

- Eliminate problems with current electromechanical designs 

• Reduce power, explosive containment, and costs 

• Make design more easy to manufacture/checkout 

• Products: 

- Flight performance demonstration-TBD 

- Guidelines for incorporating features into flight units 

- Design specification for standard safe/arm devices 
STATUS: 

• Project has been terminated 

7 - 32 - January 30. 1994 



Second NASA Aerospace Pyrotechnic Workshop 
Laser Initiated Ordnance Activities in NASA 



2.5.1 NASA STANDARD LASER DETONATOR 

Phase I - Developmental Investigations 

Project Mgr: B. Wittschen, Johnson Space Center 

• Advance pyrotechnic technology - develop laser detonators 

- Supports Project 2.4, NASA Standard Laser Diode Safe and Arm 

- Conduct off-limits testing of developmental hardware 

- Phase II task qualifies a NASA Standard Laser Detonator 

• Goals include optimizing optical interface between the fiber and the pyrotechnic charge, 
publishing a specification, and the procurement and test of devices to provide a data base 

• Products: Qualified NASA Standard Laser Detonator and design/test specification 



STATUS: 

• Project has been terminated 



January 30. 1994 



Second NASA Aerospace Pyrotechnic Workshop 
Laser Initiated Ordnance Activities in NASA 



3.4 LASER DIODE SAFE/ARM PERFORMANCE 

Project Mgr: B. Wittschen, Johnson Space Center 

• Develop test procedures 

• Quantify performance 

• Confirm specification performance 

• Demonstrate safe/arm devices for flight 

• Update design and test specifications 

• Products: 

- Publish test specification for use by programs 

- Prepare qualification report 

STATUS: 

• Project has been terminated 



9 - 33 - 



January 30, 1994 



Second NASA Aerospace Pyrotechnic Workshop 
Laser Initiated Ordnance Activities in NASA 



THEN 

RE-EVALUATE 



10 January 30, 1994 



Second NASA Aerospace Pyrotechnic Workshop 
Laser Initiated Ordnance Activities in NASA 



IMPLEMENTATION of a FEASIBILITY APPROACH: - 

BACKGROUND- 

• EVALUATED BY STEERING COMMITTEE FOR MANY YEARS 

- concern about maturity 

• AUGUST 1991: OSC/EBCO UNSOLICITED PROPOSAL TO 

CONDUCT DEMONSTRATION ABOARD PEGASUS 

- NASA performs one-time mission demonstration for a complete vehicle ordnance change 

- OSC performs fleet change 

• OBJECTIVE WAS "QUICK DEMONSTRATION" USING AVAILABLE 
TECHNOLOGY 

- delayed for two years 

- Pegasus vehicle contracted under services contract, not R&D 

- lacked clear contractual means to conduct a technology demonstration 



ii - 34 - 



January 30, 1994 



Second NASA Aerospace Pyrotechnic Workshop 
Laser Initiated Ordnance Activities in NASA 



IMPLEMENTATION APPROACH 

• MANAGERIAL ASPECTS OF LIO INITIATION POINTED TOWARD: 

- lack of technical requirements for LIO systems 

- no practical operational experience 

- lack of quick, simple, contractual instrument to implement new technology 



MANAGERIAL SOLUTION NECESSARY TO PURSUE TECHNICAL 
ISSUES 



• ABOVE ANALYSIS POINTED NEED FOR NEW LIO 
PROGRAMMATIC PATH 



12 January 30. 1994 



Second NASA Aerospace Pyrotechnic Workshop 
Laser Initiated Ordnance Activities in NASA 



STEPS REQUIRED FOR LIO IMPLEMENTATION 

1. VALIDATE FEASIBILITY 

a. ARE THE TECHNOLOGY CLAIMS CORRECT? 

b. WHAT ARE THE SAFETY, RELIABILITY PROGRAMMATIC 

DESIGN REQUIREMENTS TO FLY LASER ORDNANCE? 

IF FEASIBLE WITHIN COST COMPETITION OF EXISTING 
ELECTROMECHANICAL SYSTEMS, THEN ADDRESS THE: 

2. IMPLEMENTATION OF LIO INTO OPERATIONS 



13 - 35 - 



January .10, 1994 



Second NASA Aerospace Pyrotechnic Workshop 
Laser Initiated Ordnance Activities in NASA 



A. VALIDATE LIO FEASIBILITY: 

« REDUCE THE RISK « 

1. PERFORM FLIGHT DEMONSTRATIONS 

PHILOSOPHY: 

a. TAKE THE MANAGERIAL APPROACH OF COMMENCING WITH A MINIMUM 
SAFETY IMPACT PROJECT - THEN PROGRESS TO THE MOST DEMANDING: 

- low hazard level in a controllable application, but safety impact exists and is such that the 

LIO hazard must be controlled 

- LIO serves an active function in flight - not along just for the ride 

- ultimate application range is from unmanned to manned applications 

- ultimate system range is from flight sequencing to flight termination 

b. PERFORM SIMPLE, QUICK, DO-ABLE PROJECTS, ADDRESSING ISSUES AS 
PROGRESSION OCCURS 

2. DEVELOP REQUIREMENTS 

a. PREPARE SPECIFICATION REQUIREMENTS 

b. DEVELOP RANGE REQUIREMENTS 

14 January 30, 1994 



Second NASA Aerospace Pyrotechnic Workshop 
Laser Initiated Ordnance Activities in NASA 



B. OPERATIONAL IMPLEMENTATION 

« REMOVE THE RISK « 

1. DEVELOP A "STANDARD" 

- discussions held with Aerospace/Air Force: 

- definition of "Standard" - build to print or to performance specification 

2. QUALIFY FOR TOTAL OPERATIONAL ENVIRONMENTAL 
SPECTRUM: - CAPTURE MARKET 

3. HAVE A PRODUCT READY FOR PROGRAMMATIC USE, 
ACCEPTED BY THE PYRO TECHNICAL COMMUNITY 

4. MAINTAIN TWO QUALIFIED SOURCES AS A MINIMUM-NO SFP'S 



is - 36 - 



January 30. 1994 



Second NASA Aerospace Pyrotechnic Workshop 
Laser Initiated Ordnance Activities in NASA 



STATUS: 

THIS IS WHAT WE DID AND ARE DOING WITH 
REGARD TO THE ABOVE PROCESS 



16 January 30, 1994 



Second NASA Aerospace Pyrotechnic Workshop 
Laser Initiated Ordnance Activities in NASA 



1. PERFORM FLIGHT DEMONSTRATIONS 

a. DEVELOP A NEW PROCUREMENT PROCESS: 

COOPERATIVE AGREEMENT 
WITH PROFIT MAKING ORGANIZATIONS 



b. IMPLEMENT VIA QUICK, CHEAP FLIGHT DEMONSTRATION 
PROGRAM 



n - 37 - 



30.1994 



Second NASA Aerospace Pyrotechnic Workshop 
Laser Initiated Ordnance Activities in NASA 



COOPERATIVE AGREEMENT WITH PROFIT 
MAKING ORGANIZATIONS (CAWPMO) 

• NEW PROCUREMENT PROCESS 

- grants normally performed with universities 

- cooperative agreement previously limited by policy to non-profit organizations e.g. think-tanks, 

universities, etc. 

• CAWPMO: FROM FIRST THOUGHT UNTIL SIGNATURE = 2 MONTHS 

• FROM: RECEIPT OF PROPOSAL UNTIL SIGNATURE = 1 MONTH 

• THIS INSTRUMENT IS BASICALLY A PARTNERSHIP WITH BOTH 
GRANTEE WITH GOVERNMENT HAVING ACTIVE ROLES 

• COOPERATIVE AGREEMENT ACCOMPLISHES COMMON BENEFIT 

• NO HARDWARE IS DELIVERED 

• NO FEE 

• INTERNAL COMPANY FUNDING HELPS BUT NOT REQUIRED 

• RED TAPE REDUCED 

lg January 30, 1994 



Second NASA Aerospace Pyrotechnic Workshop 
Laser Initiated Ordnance Activities in NASA 



PROJECTS 

A. PEGASUS EXPERIMENT 

B. SOUNDING ROCKET FTS DEMONSTRATION 

C. SHUTTLE 

EFFORTS AIMED AT THE DEVELOPMENT OF REQUIREMENTS: 

- Specification 

- Range Safety 



19 - 38 - 



January 30. 1994 



Second NASA Aerospace Pyrotechnic Workshop 
Laser Initiated Ordnance Activities in NASA 



A. PEGASUS EXPERIMENT 

TWO TESTS OF LIO CONDUCTED DURING ORBCOMM MISSION: 

• CONDUCT A FLIGHT SEQUENCING FUNCTION: IGNITE TWO OF 

THE NINE FIN ROCKET MOTORS USING LIO 

- safety hazard to operational personnel: accidental motor ignition. Control by design and 

procedure 

- not mission success dependent. Fin rocket motors not required for mission success 

- qualitative information. Go-no go information. 

• FIRE LIO INTO A CLOSED BOMB 

- not a safety hazard. Accidental ignition pressurizes a metal container designed to take the load 

- not mission success dependent. Separate experiment 

- quantitative information. Pressure measurements performed during flight with be compared with 

ground test data. 

• FLY ABOARD COMMERCIAL MISSION 

- current date is June 1994 



20 January 30, 1994 



Second NASA Aerospace Pyrotechnic Workshop 
Laser Initiated Ordnance Activities in NASA 



ENSIGN BICKFORD COMPANY TASKS: 

1 . Conduct necessary design and research to demonstrate feasibility of LIO 

2. Manufacture equipment 

3 . Perform testing in coordination with NASA testing 

4. Conduct analyses 

5. Coordinate program activities closely with NASA 

6. Conduct program tasks per E-B Proposal 



21 - 39 - 



January 30, 1994 



Second NASA Aerospace Pyrotechnic Workshop 
Laser Initiated Ordnance Activities in NASA 



NASA TASKS: 

As necessary: 

1 . Perform technical review, analyses, and test support 

2. Involve Range Safety Offices 

3. Conduct off-limits/overstress tests & evaluations to support Range Safety objectives 

4. Establish requirements for NASA-wide application 

5 . Provide test equipment support such as OTDR 

6. Provide overall planning for incorporation of LIO into flight programs 

7. Conduct analyses sneak circuit analyses 

8. Conduct validation testing of sneak circuit analysis 

9. Perform FMEA, safety, and reliability analyses 

10. Conduct evaluations of program test planning 

1 1 . Conduct safety and reliability ordnance initiation evaluations 

12. Provide consultation regarding operational processes 

13. Provide guidance on generic flight operational procedures 

14. Assist in technology transfer 

22 January 30, 1994 



Second NASA Aerospace Pyrotechnic Workshop 
Laser Initiated Ordnance Activities in NASA 



B. SOUNDING ROCKET FTS DEMONSTRATION 

• OBJECTIVE: TAKE THE NEXT STEP WITH UNDERSTANDING 

REQUIREMENTS AND GAINING CONFIDENCE 

• INSTALL A FLIGHT TERMINATION SYSTEM ABOARD A TWO 

STAGE SOUNDING ROCKET AND DESTRUCT DURING THRUST 

- Nike Orion - second stage destruct flown out of Wallops 

• IGNITE FIRST AND SECOND STAGES USING LIO 

- maximize experience 

• ACTIVATE FTS BY TIMER-THIS DEMONSTRATION NOT A TEST 

TO VALIDATE NEW RF COMMAND SYSTEM 

• HIGHER LEVEL OF SAFETY REQUIRED BEYOND PEGASUS 

• 6 MONTH PROGRAM 

• AWAIT UNSOLICITED PROPOSAL FOR CAWPMO 



23 - 40 " 



January 30, 1994 



Second NASA Aerospace Pyrotechnic Workshop 
Laser Initiated Ordnance Activities in NASA 



COMPANY TASKS: 

1 . Design and manufacture termination ordnance 

2. Provide LIO ignition for Nike and Orion motors 

3. Provide laser firing unit, the fiber optic cable, connectors, detonators, and initiators 

4. Perform testing in coordination with NASA testing 

5 . Conduct analyses 

6. Install ordnance and integrate FTS/payload into launch vehicle 

7. Participate in flight operations and post flight analysis 

8. Testing at company's discretion but expected for demonstrating compatibility of 
laser initiation with current motor ignition system 



24 



January 30, 1994 



Second NASA Aerospace Pyrotechnic Workshop 
Laser Initiated Ordnance Activities in NASA 



NASA TASKS: 

1 . Launch vehicle: Nike-Orion 

2. Vehicle drawings 

3. Provide environmental test requirements and WFF range safety requirements 

4. Pyro interface such as mounting platform for LIO electronics 

5. Instrumentation defining key events and body accelerations 

6. 3-axis accelerometer 

7 . FTS activation timer 

8 . FM-FM transmitter 

9. Build-up and integration of the motor and stage assembly 

10. Flight performance analysis 

1 1 . Radar coverage 

12. Launch operations 

13. Post flight analysis support 

14. Photographic coverage 



25 - 41 " 



January 30, 1994 



Second NASA Aerospace Pyrotechnic Workshop 
Laser Initiated Ordnance Activities in NASA 



C. SHUTTLE 

• Payload (Solar Exposure to Laser Ordnance Device) 

- LIO opens shutter in space 

- Exposure of LIDS and LIS: 

- 4 different initiators 

- 2 different detonators 

- 2 different laser firing units 

- Exposure to solar radiation: 

- direct exposure to sun 

- 10:1 magnified exposure to sun 

- no exposure to sun 

- LIO subjected to Shuttle payload safety review process 

• STS Equipment (potential project not started - hazardous gas detection 

bottles) 

- Will subject LIO to Shuttle vehicle safety review process 



26 January 30, 1994 



Second NASA Aerospace Pyrotechnic Workshop 
Laser Initiated Ordnance Activities in NASA 



EFFORTS AIMED AT THE DEVELOPMENT OF 

REQUIREMENTS: 

• SPECIFICATION: UNFUNDED IN-HOUSE ACTIVITY 

• COORDINATION WITH RANGE SAFETY STAFF 

- preliminary set of requirements developed 

- work continues 



27 - 42 ~ January 30, 1994 



Second NASA Aerospace Pyrotechnic Workshop 
Laser Initiated Ordnance Activities in NASA 



PRELIMINARY 

RANGE REQUIREMENTS FOR LIOS: 

GENERAL CATEGORY "A" REQUIREMENTS: 

System Level Requirements: 

1 . Single fault tolerant (two independent safeties) before and after installation of SAFE/ ARM type connectors 

Cleared pad during power switching, power-on, and RF radiation operations 

To allow operations during these conditions, LIOS must be at least two-fault tolerant and meet the Man-Rated design 
requirements defined in RSM-93 Paragraph 5.3.4.4.5 

2 . At least one of safety controllable from pad 

3 . Design to allow power-control operations remotely from blockhouse 

4. Component (if electrical type) adjacent to the laser system must be single/double fault tolerant 

5 . Component adjacent to the lasing device (either in the power or return leg of electrical circuit), shall not be activated 
until programmed initiation event 

6. LIOS must not be susceptible to external energy sources, such as stray light energy, static and RF 

7 . Design to preclude inadvertent initiation due to singular energy sources, such as unplanned energy in power leg of 
circuit or due to short circuits or ground loops. 

8 . Design to allow for ordnance connection at the latest possible time in the countdown process 



28 January 30. 1994 



Second NASA Aerospace Pyrotechnic Workshop 
Laser Initiated Ordnance Activities in NASA 



Tri gger Circuit Requirements 

9 . Design such that voltage required to initiate laser is at least 4 times the VCC of solid state logic circuits 

10. Design to output energy after application of a 20 ms pulse 

Monitor and Test Capability: 

1 1 . Provide circuits to allow for remote control and monitor of all components in the Category "A" system. Application of 
±35V in the monitor circuit shall not affect the Category "A" circuit 

12. Recommend Built-in-Test (BIT) 

Allow remote testing at energy levels of 10" 2 below no fire for both normal and failure modes. Use different 
wavelength than main firing laser, separated by at least 100 run 

1 3. Design to allow for "no-stray energy" type of tests prior to performing ordnance connection 

14. Employ pulse catcher system to detect inadvertent actuation of laser prior to ordnance connection 

Monitor 1/100 no-fire and be capable of determining a valid all-fire (power, energy density, frequency, pulse-width) 

Laser Output Requirements; 

1 5 . Energy delivered to LID shall be 2x all-fire 

Ordnance Requirements; 

16. All ordnance used with LIOS must be secondary explosive. 



29 - 43 - 



January 30, 1994 



Second NASA Aerospace Pyrotechnic Workshop 
Laser Initiated Ordnance Activities in NASA 



Power Supply Requirements: 

1 7 . Install charged ("Hot") batteries into Category "A" circuits only if at least one of the following design approaches is 
utilized. Otherwise, charge battery at latest feasible point in countdown process with no personnel in danger area 

17.1 Electromechanical device utilized which mechanically misalign ordnance train 

17.2 Optical barriers utilized which mechanically misalign initiation power from either LID or laser 

1 7.3 Capacitive Discharge Ignition (CDI) system used meeting circuit criteria in RSM-93 Paragraph 5.3.4.4.4 

1 7.4 Designed to be Man-Rated and meets circuit requirements in RSM-93 Paragraph 5.3.4.4.5 

PRELIMINARY 
SPECIFIC CATEGORY "A" REQUIREMENTS (PARTIAL LIST): 

1 . Shielding for electrical firing circuits shall meet: 

1 . 1 Minimum of 20 dB safety margin below minimum rated function current to initiate laser and provide a minimum 

of 85% optical coverage. (A solid shield = 100% optical coverage) 
I.2Shielding shall be continuous and terminated to the shell of connectors and/or components. Electrically join 

shield to shell of connector/component around 360° of shield. Shell of connectors/components shall provide 

attenuation at least equal to that of shield 
1 .3Shield should be grounded to a single point ground at power source 
1 .40therwise, employ static bleed resistors to drain all RF power on shield 

2 . Wires should be capable of handling 150% of design load. Design shall assure that latched command will remain 
latched with a 50 ms dropout pulse 

3 . Bent pin analysis shall be performed to assure no failure modes 

4 . Analysis/Testing shall be performed to determine debris contamination for blind connection sensitivity on optical 
connectors 

5 . All components in the Category "A" initiation system shall be sealed to Kr^cc/sec 

30 January 30, 1994 



Second NASA Aerospace Pyrotechnic Workshop 
Laser Initiated Ordnance Activities in NASA 



PRELIMINARY 

FTS REQUIREMENTS: 

1 . FTS circuit must meet all requirements defined under Category "A" requirements 

2. Circuit must requirements in RCC STANDARD-319-92, FTS Commonalty Standard, Chapters 1, 2, 3, and 4 

3. All LIOS components must meet test requirements in RCC STANDARD-31992, FTS Commonalty Standard, 
Chapters 5.1 and S.2 

4 . Meet design requirements specified in WRR- 127. 1 (June 30, 1993) Chapter 4: 

a. Circuit requirements in Sections 4.6.7.4.5, 4.6.7.4.8, and 4.6.7.4.9 

b. Optical connector requirements in Section 4.7.5.2 

c. LFU requirements in Section 4.7.7.4.1 

d. LID requirements in Section 4.7.8.3 

5. System must meet test requirements specified in WRR-127.1 (June 30, 1993) Chapter 4: 

a. Appendix 4A.7: LFU Acceptance testing 

b. Appendix 4A.7: LFU Qualification testing 

c . Appendix 4A.7: Fiber Optic Cable Assembly Lot Acceptance Testing 

d . Appendix 4A.7: Fiber Optic Cable Assembly Qualification Testing 

e . Appendix 4A.7: LID Lot Acceptance Testing 

f . Appendix 4A.7- LID Qualification; Testing (need to revise numbers) 

g . Appendix 4A.7 : LID Aging Surveillance Test 
h . Appendix 4B : Common Tests Requirement 

31 - ZjZj - January 30. 1994 



Second NASA Aerospace Pyrotechnic Workshop 
Laser Initiated Ordnance Activities in NASA 



6. Incorporate a Built-in-Test (BIT) feature which allow remote testing at energy levels of 10 2 below no-fire for both 
normal and failure modes and must also be at a different wavelength than main firing laser. The wavelengths for the 
main firing laser and the test laser must be separated by at least 100 nm 

7. Piece parts shall be IAW ELV specs 

8 . All ordnance interfaces shall allow for 4 times (axial, angular max. gap) or 0. 1 5" and 50% minimum design gap 

9. Connectors per IAW MIL-C-38999J 

10. Perform analysis/design on: LIOS FTS-FMECA, bent-pin analysis, LID heat dissipation due to SPFs 



32 January 30. 1994 



Second NASA Aerospace Pyrotechnic Workshop 
Laser Initiated Ordnance Activities in NASA 



SAFETY POINTS 

• GENERAL REQUIREMENTS: 

- avoid introduction of new hazards, 

- avoid inadvertent ignition, 

- functions upon demand 

• LOW VOLTAGE FOR DIODE TO LASE CONCERN 

• POSITIVE CONTROL OF PERSONNEL SAFETY AT PAD 

ESSENTIAL 

• RANGE STRAWMAN REQUIREMENTS 



33 - 45 - 



30.1994 



Second NASA Aerospace Pyrotechnic Workshop 
Laser Initiated Ordnance Activities in NASA 



ISSUES TO WORK 

• SAFETY REQUIREMENTS 

• BUILT-IN-TEST 

• COSTS 

• DEMONSTRATED RELIABILITY UNDER VARIETY OF 
APPLICATIONS AND ENVIRONMENTS 



34 January 30, 1994 



Second NASA Aerospace Pyrotechnic Workshop 
Laser Initiated Ordnance Activities in NASA 



WHERE DO WE GO FROM HERE? 

WORK NEEDED AND NEXT STEPS 

• BUILT-IN-TEST 

• SPECIFICATION 

• DEMONSTRATED RELIABILITY UNDER A VARIETY OF 

APPLICATIONS AND ENVIRONMENTS 

• STANDARD DESIGN: BUILD TO PRINT VERSUS BUILD TO 

SPECIFICATION 

• MARKET ANALYSIS 



35 - 46 - 



January 30, 1994 



Second NASA Aerospace Pyrotechnic Workshop 
Laser Initiated Ordnance Activities in NASA 



SUMMARY 

• PROGRAMS MORE LIKELY TO USE IF CONCEPT IS PROVEN VIA 
DEMONSTRATION 



• 



• 



PROGRAMS WILL USE LIO IF QUALIFIED AND IS COST 
COMPETITIVE 

NO PROGRAM DESIRES TO MAKE USE OF LIO AND PROCEED 
DOWN THE LEARNING ROAD, UNLESS MANDATORY FOR 
PROGRAM SUCCESS OR SAFETY 

• WITHOUT A PROCESS WHEREBY THIS TECHNOLOGY IS 
DEMONSTRATED AND COST FACTORS VERIFIED, THERE IS 
NOT ANTICIPATED TO BE A DEMAND 



36 January 30, 1994 



Second NASA Aerospace Pyrotechnic Workshop 
Laser Initiated Ordnance Activities in NASA 



• 



CONCLUDING THOUGHTS 

A TECHNICAL COMMUNITY NOT UNITED IS NOT ANTICIPATED 
TO MEET WITH THE SUCCESS NECESSARY FOR LIO 
IMPLEMENTATION ON A REASONABLE TIME FRAME. 

A WELL-COORDINATED, JOINTLY-CONDUCTED, AND CO- 
FUNDED INITIATIVE BETWEEN GOVERNMENT AND INDUSTRY 
OFFERS THE BEST OPPORTUNITY FOR TECHNOLOGY 
IMPLEMENTATION. 

- One example is the Pegasus demonstration. 

- Another is laser gyro demonstration. 

THERE ARE ISSUES TO BE WORKED WITH SUCH AN APPROACH 
SUCH AS: PROPRIETARY INFORMATION, DEGREE OF FUNDING 
PARTICIPATION VERSUS RETURN EXPECTED, WHO DOES 
WHAT, GETTING AGREEMENT ON TECHNICAL ISSUES, ETC. 

BUT THESE MUST BE CONSIDERED WORKABLE WHEN VIEWED 
FROM THE PERSPECTIVE OF THE VALUE OF THE EFFORT AND 
THE IMPACT OF SUCCESS. 



37 - 47 - 



January 30, 1994 



LASER-IGNITED EXPLOSIVE AND 
PYROTECHNIC COMPONENTS 



A. C. MUNGER, T. M. BECKMAN, D. P. KRAMER 
AND E. M. SPANGLER 



AEROSPACE PYROTECHNIC SYSTEMS WORKSHOP 
FEBRUARY 8, 1994 



J^EGzG 



=4 



- 49 - 



o/V'."i 



EG&G MOUND APPLIED TECHNOLOGIES HAS PURSUED 
LASER - IGNITION TECHNOLOGIES SINCE 1980 

• EVALUATED FUNDAMENTAL EXPLOSIVE PROPERTIES 

• TESTED OPTICAL FEED-THROUGH DESIGNS 

• SHOWN FEASIBILITY OF PROTOTYPE DEVICES 

• FABRICATED & TESTED PRE-PRODUCTION QUANTITIES 



n 



EBelB 



MOUND HAS PERFORMED LASER IGNITION TESTS ON A 
VARIETY OF PYROTECHNIC AND EXPLOSIVE MATERIALS 

EXPLOSIVES PYROTECHNICS 



BARIUM STYPHNATE BCTK d 

CP° Ti/KCI0 4 

CP/CARBON BLACK TiH .65/ KCI0 4 

HMX b TiH 165 /KCI0 4 

HMX/CARBON BLACK ZrKCIOVGRAPHITE/VITON 

HMX/GRAPHITE 

HNS C 



a) 2-(5-CYANOTETRAZOLATO) PENTAAMINE COBALT (III) PERCHLORATE 

b) CYCLOTETRAMETHYLENETETRANITRAMINE 

c) HEXA-NITRO-STILBENE 

d) BORON CALCIUM CHROMATE TITANIUM POTASSIUM PERCHLORATE 



- 50 - 



COMPARISON OF THE IGNITION THRESHOLDS 
OBTAINED ON VARIOUS ENERGETIC MATERIALS 

50% ALL-FIRE IGNITION THRESHOLD 



700 mH 




100/140 STEP INDEX FIBER 
10 MS LASER DIODE PULSE 



125 mH 




75 mW 



50 mH 



CP BCTK Ti/KCIO, TiH. eg COARSE CP/ HMX/3X FINE CP/ 

4 ,„'?! IX CARBON CARBON 3. 2X CARBON 

/KL1U 4 ^CK BLACK BLACK 

ENERGETIC MATERIAL 



COMPARISON OF THE THREE PRINCIPAL DESIGNS 
UNDER CONSIDERATION FOR LASER-IGNITED COMPONENTS 



(1) 



CHARGE CAVITY 



IS 



SHELL ^V 



'i 



^m^ 



OPTICAL FIBER 



(2) 



CHARGE CAVITY 



SHELL S> 



J 



R3 



iS]^ 



FIBER PIN 



/ ^ x CHARGE CAVITY 

(3) IS I 153 

SH " 1 i 



^n^ 



WINDOW 



- 51 - 



COMPARISON OF THE 50% ALL-FIRE THRESHOLDS 
USING SAPPHIRE AND P-GLASS WINDOW DEVICES 



"3 

o 
lu 



4.0-1 



3.0- 



2.0- 



w 1.0- 



0.0 



SAPPHIRE 



P-GLASS 






1 1 1 — 

0.0 0.2 0.4 0.6 

WINDOW THICKNESS (mm) 



MOUND HAS DESIGNED AND FABRICATED OVER 
TEN DIFFERENT LASER-IGNITED COMPONENTS 



• 3 "FIBER PIGTAIL" PROTOTYPES 

• 5 "WINDOW" 

• 6 "FIBER PIN" 

• LOT SIZES - UP TO 400 COMPONENTS 



&> 



- 52 



MOUND IS ACTIVELY ENGAGED IN LASER DIODE 
IGNITION (LDI) COMPONENT DEVELOPMENT 






SEALED 


SEALED 


HIGH 


WINDOW 


FIBER 


STRENGTH 


DETONATOR 


DETONATOR 


ACTUATOR 



^O 



EGkG 



HERMETIC, LASER-IGNITED DEFLAGRATION 
TO DETONATION TRANSITION (DDT) DETONATOR 




- 53 - 



IGNITION THRESHOLD DATA ACQUISITION SYSTEM 



FIBER FOCUSING SHUTTER 

COUPLER LENS 





BEAM CHOPPER 

EXPANDER 



CW Nd:YAG LASER 



BARRICADE WfTH OPTICAL WINDOW 



STC CONNECTER 



OPTICAL 
FIBER 



HIGH SENSITIVITY 
PHOTODIODE 






w& 



-^ 



PHOTODIODE 



TEST DEVICE 






«a» 



DIGITIZING OSCILLOSCOPE 



DlGmZING OSCILLOSCOPE 



IPS92-10 



SEVERAL PARAMETERS MUST BE CONSIDERED WHEN 
REFERRING TO COMPONENT THRESHOLD VALUES 

• LASER BEAM/POWDER INTERFACE 

- SPOT SIZE 

- SPOT SHAPE 

• LASER BEAM 

- PULSE LENGTH 

- PULSE SHAPE 

- WAVELENGTH 

• THERMAL CONDUCTIVITY 

- POWDER/FIBER/SHELL/WINDOW 



- 54 - 



THE DDT DETONATOR HAS BEEN SUCCESSFULLY 
LASER-FIRED USING A VARIETY OF TEST PARAMETERS 



Maximum 50% All-Fire Standard 
Laser Fiber Dia. N.A. Pulse Length Threshold Deviation 



Nd:YAG 


1000(1 


0.22 


12 msec 


30 mJ 


8 mJ 


Nd:YAG 


200|i 


0.37 


150 i^sec 


34 mJ 


16 mJ 


Nd:YAG 


200^ 


0.22 


12 msec 


50 mJ 


11 mJ 



A 



ONE OF THE SEALED FIBER DEVICES HAS BEEN 
SUCCESSFULLY CHARACTERIZED 



• HMX/CARBON BLACK 

• HERMETIC 

• 50% ALL-FIRE IGNITION 
THRESHOLD ~1.4mJ 

• MAINTAINED STRUCTURAL INTEGRITY 
MAXIMUM PRESSURE -20,000 psi 



A 




HMX 



OPTICAL FIBER 



HMX LASER SQUIB 



- 55 - 



The Laser Fired Pyrotechnic device 

was derived from the well tested 

" Hot-Wire" Device 



ELECTRIC PYROTECHNIC SQUIB 




TiH /KCIO. 
1 1.65 4 

BRIDGE WIRE 
RTVPAD 



LASER PYROTECHNIC SQUIB 



A 




™ i-66 «cio 4 



EXAMPLE OF A HIGH-STRENGTH LASER 
IGNITED PYROTECHNIC ACTUATOR 






■ ■».-» 



fH'-r'iifM 1 ' 

im 10 2 
i.l.ilfllli. 



Conversion Rule ?f 




.iitlUtJtiillllt.lt 



- 56 - 



THRESHOLD PERFORMANCE INDICATES THAT A 
RELIABLE DEVICE CAN BE FABRICATED 

THRESHOLD DATA WITH 10 ms PULSE 



ENVIRONMENT 

NONE 

TS.TC 
TC.TS 

NONE, MYLAR 
TC, MYLAR 



ENERGY 


TEMP C 


5.3 mJ (0.05) 


•55 


5.02 mJ (0.7) 
4.50 mJ (0.2) 


•55 
-55 


3.3 mJ(1.2) 
2.7 mJ (0.5) 


-55 
-55 



&** 



ZERO VOLUME FIRING TEST SETUP USED TO DETERMINE 
PRESSURE OUTPUT OF LASER-IGNITED FIBER PIN DEVICE 



DATA ACQUISITION 



DIODE DRIVER 



-0 



OPTICAL 
FIBER \ 



*r 



LASER 
DIODE 



TRANSDUCER 



FIBER PIN 
j DEVICE \ 



»»»>»jjjiit>*i»\>iii*j*\jjjt 




PRESSURE BLOCK 






""""TV. 



EXPLOSIVE TEST CHAMBER 



-57 - 



ZERO VOLUME FIRING TEST RESULT OBTAINED 
ON A LASER-IGNITED FIBER PIN DEVICE 



1200 



174 




120 160 
TIME (microsecond) 



ii 



COMPACT" RIGHT-ANGLE LASER-IGNITED 
DEVICES CAN BE MANUFACTURED 




fcNIIIIM 

in xVio , I 



im I 

rfffii. 



o 

,iii 



M I | | I I \ M M I \ \ 

Conversion Rule 2| 



2 3 4 5I0 6< 



6\0 

\ 



- 58 



LASER-IGNITED DETONATORS HAVE BEEN 
DESIGNED TO BE BON-FIRE SAFE 

• TWO PROTOTYPE STYLES HAVE BEEN TESTED 



TESTS HAVE SHOWN THAT DETONATION DID NOT 
OCCUR DURING THERMAL EXCURSION 



DEVICE WILL NOT RECOVER 



J^EB&B 



LDI COMPONENTS HAVE BEEN FABRICATED 
AT MOUND TO SUPPORT A VARIETY OF TESTING 

MOUND TESTING 

THRESHOLD DETERMINATION 
HOT, COLD, AMBIENT 

ENVIRONMENTAL CONDITIONING 
THERMAL AND MECHANICAL 

KED TESTING ( ZERO VOLUME ) 
VALVE ACTUATOR 

SANDIA TESTING 
LIGHTNING STRIKE 

ESD - FISHER MODEL 

- SANDIA STANDARD MAN 

- 59 - 



LASER DETONATOR MANUFACTURING REQUIRES 
THE APPLICATION OF SEVERAL KEY TECHNOLOGIES 



GLASS PROCESSING 



NONDESTRUCTIVE 
EVALUATION 



FURNACE 
TECHNOLOGY 

GLASS MACHINING 



LASER DDT- 
DETONATOR- 



/ \ 



LASER WELDING 
GLASS SEALING 
- POWDER PRESSING 
HEAT TREATING 



EXPLOSIVE POWDER 
PROCESSING 



- 60 - 



1994 MhBA PYROTEOHNI© SYSTG 






A LOW COST IGNITER UTILIZING AN SCB AND TITANIUM 
SUB-HYDRIDE POTASSIUM PERCHLORATE PYROTECHNIC 



R. W. Bickes, Jr. and M. C. Grubelich 

Sandia National Laboratories 

Albuquerque, NM 87185-0326 

J. K. Hartman and C. B. McCampbell 

SCB Technologies, Inc. 

Albuquerque, NM 87106 

J. K. Churchill 

Quantic-Holex 

Hollister, CA 



ABSTRACT 

A conventional NSI (NASA Standard Initiator) normally employs a hot-wire ignition element to ignite ZPP 
(zirconium potassium perchlorate). With minor modifications to the interior of a header similar to an NSI 
device to accommodate an SCB (semiconductor bridge), a low cost initiator was obtained. In addition, the 
ZPP was replaced with THKP (titanium subhydride potassium perchlorate) to obtain increased overall gas 
production and reduced static-charge sensitivity. This paper reports on the all-fire and no-fire levels 
obtained and on a dual mix device that uses THKP as the igniter mix and a thermite as the output mix. 



ll^lii;: 



1 . INTRODUCTION 



The Explosive Components Department at 
Sandia National laboratories was assigned the 
task of designing actuators for several different 
functions for a Department of Energy (DOE) 
program. The actuators will be exposed to 
personnel as well as to a wide variety of 
mechanical, temperature and electromagnetic 
environments. In addition, required outputs vary 
from a high pressure gas pulse for piston 
actuation to a high temperature thermal output 
for propellant ignition. In order to minimize 
complexity, the firing sets for all the actuators 
must be the same, and the firing signal must be 
transmitted via a cable over lengths as long as 
thirty feet. 



Our solution was to modify an existing Quantic- 
Holex component (similar to a conventional NSI 
device) with a semiconductor bridge, SCB. Our 
prototype device used titanium subhydride 
potassium perchlorate (THKP) as the 
pyrotechnic. Our second (dual mix) design used 
THKP as the igniter mix and CuO/AI thermite as 
the output charge. The low firing energy 
requirements of the SCB substantially reduced 
the demands on the firing system; indeed, the 
present firing system design could not 
accommodate conventional hot-wire devices. 
The reduced static sensitivity of THKP 1 helped 
mitigate the electromagnetic environment 
requirements for exposure to radio frequency 



This work performed at Sandia National Laboratories is supported by the U. S. Department of Energy 
under contract DE-AC04-76DP00789. Approved for public release; distribution unlimited. 



- 61 - 



(RF) signals and 
discharges (ESD). 



human-body electrostatic 



2. SCB DESCRIPTION 

An SCB is a heavily doped polysilicon volume 
approximately 100 urn long by 380 urn wide and 
2 |im thick with a nominal resistance of 1 Q. It is 
formed out of the polysilicon layer on a 
polysilicon-on-silicon wafer. Aluminum lands are 
defined over the doped polysilicon; wires are 
bonded onto the lands connecting the lands to 
the electrical feed-throughs of the explosive 
header. The firing signal is a short (30 lis) current 
pulse that flows from land-to-land through the 
bridge. The current melts and vaporizes the 
bridge producing a bright plasma discharge that 
quickly ignites the THKP pressed against the 
bridge. 2 




Figure 1. Simplified sketch of an SCB. 



The main advantages of an SCB igniter versus 
conventional hot-wire igniters are that (1) the 
input energy required to obtain powder ignition 
is a decade less than for hot wires; (2) the no-fire 
levels are improved due to the large heat sinking 
of the silicon substrate; and (3) the function 
times (i.e. the time from the onset of the firing 
pulse to the explosive output of the devices) are 
only a few tens of microseconds or less. 3 



3. IGNITER DESIGN 

Our SCB igniter is similar to the NSI device. It 
consists of a metal body containing a glass 
header, charge holder and pyrotechnic material. 
3/8-24 UNF threads allow the device to be 
installed into test hardware and an O-ring under 
the hexagonal head provides the gas seal. The 



internal charge cavity was reduced to a diameter 
of 0.156" by utilizing a threaded fiber glass 
composite (G10) charge holder. The threads 
help prevent separation of the powder from the 
bridge due to mechanical shock. The pins are 
hermetically sealed by glass-to-metal seals and 
extend approximately 0.020" past the header 
base into the charge holder. The SCB chip is 
bonded to the header base between the pins 
with a thermally conductive epoxy. Aluminum 
wires, 0.005" in diameter, are thermalsonically 
bonded to the header pins and the aluminum 
lands on the chip. For the prototype device, a 
charge of 85 mg of THKP is pressed at 12,500 
psi into the charge holder. A G10 disk, 0.156" 
diameter and 0.010" thick, is placed on top of the 
pressed powder, followed by an RTV disk, 
0.150" diameter and 0.016" thick. A G10 plug, 
0.065" thick, is then pressed on top of the RTV 
at a pressure sufficient to compress the RTV pad 
to half its thickness, which maintains a pressure 
of approximately 5,000 psi on top of the THKP 
column. A high temperature epoxy seals the 
interference fit G10 plug in place. 



CLASS-TO- METAL SEAL 

IGNITER BODY. 304L CRES 
IX PM. 302 ORES 




C- ID PHENOLIC SUEVE 

OUTPUT CLOSURE. WELDED 



J7ft-24liMF-2A 



Figure 2. The SCB igniter outline. 



4. FIRING SET DESCRIPTION 

The firing set for our application is low voltage 
capacitor discharge unit (CDU) with a 50 \iF 
capacitor charged to 28 V (nominal). 4 Because 
the SCB dynamic impedance changes 
significantly during the process that produces 
the plasma discharge, two FET switches in 
parallel are required to discharge the 35 A 
current pulse into the SCB. In addition a test 
current pulse is included that passes a 10 mA 
pulse through the bridge to verify igniter 
integrity. 



- 62 - 



WW IN 14-J1V • 



i — | mutis 'f twit 



n 



-rn Wi 

n 



Figure 3. Wiring schematic for the SCB low voltage 
CDU firing set. 



As noted in section 1. ( some of the igniters may 
be located as far as thirty feet from the firing set. 
The use of either large diameter wire pairs or 
ordinary BNC cable reduced the transmitted 
current pulses to levels below threshold for 
ignition. However, Reynolds Industries "C" cable 
was able to transmit the current pulse with only a 
small attenuation of the peak current. 



5. PROTOTYPE TESTS 

Figure 4. shows the voltage, current and 

40. 



O 30. 
X 

O 

c 

CO 

O £0. 

Q. 

£ 

I 

E 
| u>. 



0.0 



/ ' \ 


v." 


A V I 


' 


' Z 


1 



1.0 2.0 



3.0 4.0 

time (us) 



5*0 



6.0 



7.0 



Figure 4. Current (I), voltage (V) and impedance (Z) 
wave forms across the SCB. At 4.8 //s the peak 
current was 37.4 A, the corresponding voltage was 
19. 1 V and the impedance 0.5 Q. 
impedance wave forms across a device fired 
when connected to the firing set through 30 feet 
of "C" cable. At ambient conditions, the device 
functioned in 83 |is (determined by a 



photomultiplier tube looking at the end of the 
device thorough a fiber optic cable). 



Ten units were tested using the NEYER/SENSIT 
scheme. 5 The units were fired at ambient and 
connected to the firing set with 30 feet of "C" 
cable. AnASENT 6 analysis of the data indicated 
a mean all-fire voltage of 17.8 V + 0.2 V; 
confidence limits on the mean were 17.7 to 19.2 
V at a 95% confidence level and a probability of 
function of 0.999. See table I for a listing of the 
data in the shot order prescribed by SENSIT. 



TABLE I: ALL-FIRE DATA 



Cap Voltage 


Go/Noqo 


Energy 


(V) 


(X/O) 


(mJ) 


18.0 


X 


5.01 


17.0 





4.36 


17.5 





4.63 


18.2 


X 


4.63 


17.7 


X 


4.77 


17.2 





4.54 


17.9 





4.90 


17.3 





4.54 


18.2 


X 


5.08 


17.4 





4.66 



All Fire: 17.8±0.2 V, 4.8±0.1 mJ 



Six THKP units underwent 3 temperature cycles 
over a twenty-four hour period. Each cycle 
consisted of 4 hours a 74C and 4 hours at -54C. 
The devices were fired as soon as possible after 
the cold cycle at approximately -15C. All of the 
units function when fired using the firing set 
without the 30 foot cable. 

We subjected a THKP unit to a 1 A Current for 5 
minutes. There was no indication of device 
degradation and the unit functioned properly 
when tested. Based on the no-fire tests in Ref. 
3, which used the same bridge as tested in this 
paper, we are confident that these units will have 
similar no-fire levels similar to those reported in 
Ref. 3(1.39±0.03A). 



6. DUAL MIX DEVICE 

Because composite propellants require a 
relatively large amplitude long duration thermal 



- 63 - 



input for reliable ignition, we developed an SCB 
igniter employing two discrete pyrotechnic 
compositions. First, 25 mg of THKP is pressed at 
12.5 kpsi against the SCB and is used as a starter 
mix to pyrotechnically amplify the low energy 
SCB signal. The THKP in turn ignites and ejects 
150 mg of a high density thermite composition 
composed of CuO and Al pressed onto the 
THKP. 

We briefly describe the advantages of this device 
over a device composed of only a single load of 
THKP or CuO/AI. THKP has excellent and well 
known interface, ignition and pyrotechnic 
propagation properties. It also is an excellent gas 
producer providing zero volume pressures 
greater than 150 kpsi. Unfortunately, the short, 
high pressure output pulse of THKP is not ideally 
suited for the ignition of a composite propellant. 
CuO/AI on the other hand is an ideal material for 
the ignition of composite propellants. Hot 
copper vapor condensing and molten copper 
impacting on the surface of the propellant 
provides an excellent source of thermal energy 
for ignition. Furthermore, copper and copper 
oxides catalyticaily enhance the ignition and 
combustion of ammonium perchlorate. 
Unfortunately, CuO/AI thermites exhibit poor 
ignition characteristics at high density and are 
sensitive to header and charge holder thermal 
losses. Thus, CuO/AI at high density requires 
large input energies for ignition and the reaction 
once started can be quenched as a result of 
radial heat losses. The THKP ignition charge 
eliminates both of these problems by providing 
an overwhelming thermal input to the CuO/AI. 
Although the CuO/AI is itself a poor gas producer 
(the copper vapor rapidly condenses), the THKP 
produces a sufficient gas pulse for this device to 
be used to operate small, lightly loaded, piston 
type actuators. In addition, the thermal output of 
the CuO/AI helps to maintain the temperature of 
the gases produced by the THKP. We have 
tested both piston actuator and propellant 
loaded gas generators with this dual mix device 
with good results. 



using a 50 \xF CDU firing set was 17.8 V; the 5 
minute no-fire level is estimated to be greater 
than 1 A with no device degradation. Future 
research will examine the tolerance of this device 
to mechanical shock and electromagnetic 
environments. 



8. ACKNOWLEDGMENT 

The testing expertise of Dave Wackerbarth, 
Sandia National Labs, is acknowledged with 
grateful thanks. 



9. REFERENCES 

1 E. A. Kjelgaard, "Development of a Spark 
Insensitive Actuator/Igniter," Fifth International 
Pyrotechnics Seminar, Vail Colorado (July 
1976). 

2 See for example, D. A. Benson, M. E. Larson, 
A. M. Renlund, W. M. Trott and R. W. Bickes, Jr., 
"Semiconductor Bridge (SCB): A Plasma 
Generator for the Ignition of Explosives," Journ. 
Appl. Phys. 62, 1622(1987) 

3 R. W. Bickes, Jr., S. L. Schlobohm and D. W. 
Ewick, "Semiconductor Bridge (SCB) Igniter 
Studies: I. Comparison of SCB and Hot-Wire 
Pyrotechnic Actuators," Thirteenth 
International Pyrotechnic Seminar, Grand 
Junction Colorado (July 1988). 

4 Firing set designed by J. H. Weinlein of the 
Firing Set and Mechanical Design Department, 
Sandia National Laboratories. 

5 B. T. Neyer, "More Efficient Sensitivity Testing," 
EG&G Mound Applied Technologies, MLM- 
3609, (October 20, 1989) 

6 H. E. Anderson, "STATLIB," Sandia National 
Laboratories, SAND82-1976, (September 
1982). 



7. SUMMARY 

We have developed two SCB igniters housed in 
an assembly with an outline similar to the 
standard NSI component. Our prototype design 
utilized THKP to provide for a pressure output 
static-insensitive device. Our second design 
used a THKP and thermite mix to provide an 
output sufficient for piston actuators as well as 
propellant loaded gas generators. All-fire voltage 



- 64 - 



b 



ft •'« 



Optical Ordnance System For Use In Explosive Ordnance Disposal Activities* 



J. A. Merson, F. J. Salas, and F. M. Helsel 

Explosives Subsystems and Materials Department 2652 

P. O. Box 5800 

Sandia National Laboratories 

Albuquerque, NM 87185-0329 



ABSTRACT 

A portable hand-held solid state rod laser system and 
an optically-ignited detonator have been developed 
for use in explosive ordnance disposal (EOD) 
activities. Laser prototypes from Whittaker 
Ordnance and Universal Propulsion have been tested 
and evaluated . The optical detonator contains 2-(5 
cyanotetrazolato) pentaamine cobalt III perchlorate 
(CP) as the DDT column and the explosive 
Octahydro - 1,3,5,7 - tetranitro - 1,3,5,7 - tetrazocine 
(HMX) as the output charge. The laser is designed 
to have an output of 150 mJ in a 500 microsecond 
pulse. This output allows firing through 2000 
meters of optical fiber. The detonator can also be 
ignited with a portable laser diode source through a 
shorter length of fiber. 



1.0 INTRODUCTION 

Sandia National Laboratories has been actively 
pursuing the development of optically ignited 
explosive subsystems for several years concentrating 
on developing the technology through experiment 1 ' 3 
and numerical modeling of optical ignition. 4 * 5 
Several other references dealing with various aspects 
of optical ordnance development are also available in 
the literature. 6 " 1 ^ Our primary motivation for this 
development effort is one of safety, specifically 
reducing the potential of device premature that can 
occur with a low energy electrically ignited explosive 
device (EED). Using laser ignition of the energetic 
material provides the opportunity to remove the 
bridgewire and electrically conductive pins from the 



charge cavity, thus isolating the explosive from stray 
electrical ignition sources such as electrostatic 
discharge (ESD) or electromagnetic radiation 
(EMR). The insensitivity of the explosive devices to 
stray ignition sources allows the use of these 
ordnance systems in environments where EED use is 
a safety risk. 

The Office of Special Technologies under the 
EOD/LIC program directed the development of a 
portable hand-held solid state rod laser system and 
an optically-ignited detonator to be used as a 
replacement of electric blasting caps for initiating 
Comp C-4 explosive or detonation cord in explosive 
ordnance disposal (EOD) activities. The prototype 
systems that have been tested are discussed in this 
paper. Laser prototypes were procured from both 
Whittaker Ordnance (now Quantic) and Universal 
Propulsion Company and tests were conducted at 
Sandia National Laboratories. An optical detonator 
was designed at Sandia National Laboratories and 
built by Pacific Scientific - Energy Dynamics 
Division formerly Unidynamics in Phoenix (UPI). 



2.0 THEORY OF OPERATION 

The intent of the optical firing system is to provide 
the same functional output performance of an 
electrically fired blasting cap without the use of 
primary explosives. Electrical detonation systems 
use current to heat a bridgewire which in turn heats 
an explosive powder to its auto-ignition temperature 
through conduction. In contrast, an optical system 
uses light energy from a laser source that is absorbed 



♦This work was sponsored by the Office of Special Technologies under funding documents NO464A92WR07053 
and NO464A91WR10380 and supported by the United States Department of Energy under Contracts DE-AC04- 
76DP00789 and DE-ACO4-94AL85000. 



- 65 - 



by the powder, thus raising its temperature to the 
auto-ignition temperature. The primary advantage 
of optical ignition is that there are no electrically 
conductive bridgewires and pins in direct contact 
with the explosive powder. This removes the 
potential electrostatic discharge pathways and 
eliminates premature initiations which can be caused 
by stray electrical signals. This is illustrated by the 
comparison of the electrically and optically ignited 
ordnance systems shown in Figure 1. 



ELECTRICAL 
DEVICE 



BRIDGEWIRE 



POWER 
SOURCE 



ELECTRICAL 
LEADS 



y^OWDE^\ 




W 



HERMETIC WINDOW 
OR FIBER OPTIC 



OPTICAL 
FIBER 




standard SMA 906 optical connector. The 
connector positions the optical fiber in contact with a 
sapphire window as shown in Figure 2. This optical 
interface and the use of optical fibers instead of 
electrical wires completely de-couples stray electrical 
sources from the detonator by removing any 
electrical path to the explosive. 




E-BRfTE 26-1 



HMX Output Chwg* 
CP DDT Column 1.7 oVce 
1.5 glee 



H 



Figure 2. SMA compatible optical detonator with 

doped CP ignition charge, undoped CP DDT column 

and a HMX output charge. 



Figure 1. Comparison of electrically and optically 
ignited ordnance systems. 



3.0 SYSTEM DESCRIPTION 

The optical system is intended to be an additional 
tool for EOD applications which provides a HERO 
(Hazards of Electromagnetic Radiation to Ordnance) 
safe system with a detonation output sufficient to 
directly initiate Comp C-4 or detonation cord 
without the use of primary explosives such as Lead 
Azide. The system contains an optical detonator, a 
portable, battery operated laser, and optical fiber to 
couple the laser output to the detonator. Each part of 
the system will be discussed individually. 

3.1 Detonator Description 

A drawing of the detonator design is shown in 
Figure 2. The detonator relies upon the deflagration 
to detonation transition or DDT. The detonator 
contains approximately 90 mg of 2-(5- 
cyanotetrazolato) pentaamine cobalt III perchlorate 
or CP (see Figure 3 for chemical structure) for the 
DDT column and 1 g of HMX for the output charge. 
The detonator wall around the HMX output charge is 
thin in order to minimize the attenuation of the 
shock produced by the detonation of the HMX. The 
detonator incorporates threads that will accept a 



N C-CN 

II II 

N N 

\ / 

N 



H J 



H ^ 



Y±7 

NH 



NH 



(CIO 



4' 2 



Figure 3. 2-(5-cyanotetrazolato) pentaamine cobalt 
III perchlorate or CP. 

The optical ignition of explosives depends on the 
optical power delivered and the energy absorbed by 
the explosive. This dependence is important at low 
power as shown Figure 4. 

At low power, it is necessaiy to dope some 
explosives with other materials such as carbon black 
or graphite in order to increase their absorptance of 
the optical energy and thus lower their ignition 
threshold. We have chosen to use CP doped with 
1% carbon black so that these detonators can be fired 
from lower power laser sources such as laser diodes. 



- 66 - 



At high powers, such as that provided by the Navy 
EOD system, a minimum energy must be delivered 
to the explosive in order for it to ignite. As seen in 
Figure 4, this minimum energy for doped CP is on 
the order of 0.25 mJ. The Navy EOD system uses a 
solid state rod laser capable of delivering 100 to 200 
mJ of optical energy in a fraction of a millisecond. 
Explosive doping is not required in this detonator 
when utilizing the high power rod laser but was 
implemented so that the detonator could be used for 
a wide range of applications. 



30,000 



Energy (mJ) 




0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 

Power (watts) 

Figure 4. Optical ignition threshold for doped CP at 
low laser powers. 

Successful ignition and function of the optical 
detonator has been achieved with both portable solid 
state rod laser systems powered by a 9 V supply and 
by a portable semiconductor laser diode system 
powered by five 9 V batteries (45 V total). The 
operational goals of the detonation system require 
the use of long optical fiber lengths (up to 2000 m) 
which may have optical attenuation or loss near 90 
percent with fibers that have 4-5 dB/km loss. 
Fibers with higher loss per kilometer will enhance 
the optical attenuation problem. The portable laser 
diode is capable of delivering 2 W of optical power, 
well within the ignition requirements, but 
insufficient to overcome the cable losses in 2000 m 
of optical cable. For this reason, the EOD system 
uses solid state rods for the optical energy supply 
which are discussed in the next section. 



3.2 Solid State Rod Laser 

Two laser firing unit designs have been built by 
Whittaker Ordnance (now Quantic) and by Universal 
Propulsion Company in Phoenix. The Whittaker 
design was the first generation prototype followed by 



the second generation prototype design from 
Universal Propulsion. Both systems have been 
shown to be effective at igniting the optical detonator 
through 1000 meters of optical fiber. Both laser 
designs are discussed below. 

The first laser firing unit for the Navy EOD laser 
ordnance system was built by Whittaker Ordnance 
and was designed to be portable, rugged, water-proof 
during transport, and battery operated. The laser 
unit contains a 9-volt battery which supplies voltage 
to a DC/DC converter to step up the voltage to 
approximately 500 volts. This voltage charges the 
300 nf capacitor which supplies current to the flash 
lamps. The functioning of the flashlamps excites the 
laser rod material, Nd doped YAG, and causes the 
laser to function. The system is designed to deliver 
between 100 and 200 mJ of optical energy during a 
500 microsecond pulse. This exceeds the energy 
required for the ignition of the detonator by at least 2 
orders of magnitude. The laser output is coupled 
into a 200 urn optical fiber which can be connected 
to the laser firing unit using the SMA 905 connector 
port on the top of the laser. 

The laser can be easily transported in the field. It is 
contained in a cylindrical container which is 
approximately 3.5 inches in diameter and 6 inches 
tall. The package weighs about 2 pounds. The laser 
is not eye safe and care must be taken to properly 
protect the operator and any casuals from exposure 
to the beam. Laser safety glasses with an optical 
density of 4.6 or greater are required for personnel 
within 10 feet or 3 meters of the laser or the output 
end of a fiber when it is coupled to the laser. During 
operations, one person maintains positive control of 
the laser and the optical detonators. It is the 
responsibility of that person to assure that all 
personnel within the exposure radius of 3 meters 
have the proper eye protection. Once this is verified, 
the laser can be armed by depressing the arm button 
on the top of the laser firing unit. After 10 to 30 
seconds, the fire light will begin to blink. The laser 
can then be fired by depressing the fire button. The 
optical fiber can then be disconnected from the laser 
and the protective cover placed back on the optical 
port on the laser. 

The second generation laser was designed and built 
by Universal Propulsion Company. It improved 
upon the packaging, specifically with respect to 
environmental protection, and maintained a 
comparable laser output to the Whittaker laser. This 



- 67 - 



laser uses either six 1.5 V AA batteries or three 3 V 
AA batteries to power the laser with a 9 V supply. 
The 9 V supply is stepped up to 360 V to charge a 
200 |if capacitor. The body of the laser is more 
rugged and environmentally sealed. The housing is 
similar to a flashlight housing and is 10.1 inches 
long and 2.75 inches in diameter. The laser weighs 
2.1 pounds. Operation of the laser is similar to that 
of the Whittaker design. The design utilizes a rotary 
arm/fire switch located in the rear of the laser 
housing. The laser delivers 200 - 300 mJ optical 
energy in a 200 usee pulse. The optical energy is 
coupled into a 200 jam fiber using a press fit SMA 
906 connector which attaches to the front of the laser 
housing. 



3.3 Optical Fiber and Connections 

The optical energy from the laser is coupled to the 
optical detonator with the use of optical fiber. The 
fiber contains a core glass and either a glass or 
plastic cladding depending on the manufacturer. 
The mismatch of the index of refraction of the core 
and cladding is such that all of the optical energy in 
the core glass is internally reflected by the cladding 
in a process known as total internal reflectance. 
Each optical fiber is described by a size and 
numerical aperture (NA). The size of the fiber is 
determined by the core glass diameter. The Navy 
EOD system uses 200 urn fiber and could easily be 
adapted to larger diameters such as 400 urn. The 
NA of the fiber describes the acceptance angle of the 
light that can be coupled into the fiber such that the 
light in the fiber does not exceed the critical angle 
and is totally internally reflected. The core and 
cladding are coated with an organic buffer to add 
strength. Additional layers of plastic and other 
strength members including Kevlar are used in the 
optical fiber cable to give it additional strength. The 
overall cable diameter can vary depending upon the 
jacketing and strength member materials but is on 
the order of 0.125 inches. 

The optical fiber is relatively durable, however it can 
be broken. Care should be taken to avoid sharp 
bends less than 0.5 inch radius. Using a visible light 
source which should be eye safe, the operator can 
check for breaks in the optical cable by shining the 
light through the fiber. During system setup, the 
light can be transmitted through the fiber to verify 
continuity. If the light does not appear at the other 
end, then there is a break in the fiber cable. Only an 



eye safe, low power, light source should be used for 
checking fiber continuity. The fiber continuity 
cannot be checked by the laser firing unit as it is not 
eye safe, and the laser light is invisible to the human 
eye. 

Connections to optical fibers can be made with 
standard optical connectors. This procedure can be 
done in the field if required but is easier if done 
ahead of time. The polish on the optical fiber is 
important on the laser end. The polish on the 
detonator end is not as critical and a simple cleave of 
the fiber is sufficient. During explosive shots, the 
last portion of the optical fiber is destroyed. 
Therefore, it is recommended that optical cable 
jumpers be prepared ahead of time and used in the 
field to minimize the number of connectors that are 
made in the field. 



4.0 SUMMARY 

The optical ordnance system utilizes laser light 
energy to ignite an explosive powder contained in a 
detonator. The detonator is HERO safe and 
produces a detonation output sufficient to detonate 
Comp C-4 or detonation cord. The detonator does 
not contain primary explosives. The laser is portable 
and powered by batteries. The optical energy from 
the laser is coupled into standard optical fiber which 
is connected to the detonator. Jumpers are used to 
minimize the number of optical fiber terminations 
that must be made in the field with multiple shots. 
The system has been shown to be effective at 
detonating Comp C-4 through 1000 meters of optical 
fiber. 



5,0 REFERENCES 

1. S. C. Kunz and F. J. Salas, "Diode Laser 
Ignition of High Explosives and Pyrotechnics", 
Proceedings of the Thirteenth International 
Pyrotechnics Seminar, Grand Junction, CO, 
11-15 July 1988, p. 505. 

2. R. G. Jungst, F. J. Salas, R. D. Watkins and L. 
Kovacic, "Development of Diode Laser-Ignited 
Pyrotechnic and Explosive Components", 
Proceedings of the Fifteenth International 
Pyrotechnics Seminar, Boulder, CO, 9-13 July 
1990, p. 549. 

3. J. A. Merson, F. J. Salas and J. G. Harlan, "The 
Development of Laser Ignited Deflagration-to- 
Detonation Transition (DDT) Detonators and 



68 - 



Pyrotechnic Actuators", to be published in 
Proceedings of the Nineteenth International 
Pyrotechnics Seminar, Christchurch, New 
Zealand, 20-25 February, 1994. 

4. M. W. Glass, J. A. Merson, and F. J. Salas, 
"Modeling Low Energy Laser Ignition of 
Explosive and Pyrotechnic Powders", 
Proceedings of the Eighteenth International 
Pyrotechnics Seminar, Breckenridge, CO, 12- 
17 July 1992, p. 321. 

5. R. D. Skocypec, A. R. Mahoney, M. W. Glass, 
R. G. Jungst, N. A. Evans and K. L. Erickson, 
"Modeling Laser Ignition of Explosives and 
Pyrotechnics: Effects and Characterization of 
Radiative Transfer", Proceedings of the 
Fifteenth International Pyrotechnics Seminar, 
Boulder, CO, 9-13 July 1990, p. 877. 

6. D. W. Ewick, "Improved 2-D Finite Difference 
Model for Laser Diode Ignited Components", 
Proceedings of the Eighteenth International 
Pyrotechnics Seminar, Breckenridge, CO, 12- 
17 July 1992, p. 255. 

1. D. W. Ewick, T. M. Beckman, J. A. Holy and 
R. Thorpe, "Ignition of HMX Using Low 
Energy Laser Diodes", Proceedings of the 
Fourteenth Symposium on Explosives and 
Pyrotechnics, Philadelphia, PA, 1990, p. 2-1. 

8. D. W. Ewick, T. M. Beckman and D. P. 
Kramer, "Feasibility of a Laser-Ignited HMX 
Deflagration-to-Detonation Device for the U. S. 
Navy LITES Program", Rep. No. MLM-3691, 
EG&G Mound Applied Technologies, 
Miamisburg, OH, June 1991, 21 pp. 

9. C. M. Woods, E. M. Spangler, T. M. Beckman 
and D. P. Kramer, "Development of a Laser- 
Ignited All-Secondary Explosive DDT 
Detonator", Proceedings of the Eighteenth 
International Pyrotechnics Seminar, 
Breckenridge, CO, 12-17 July 1992, p. 973. 

10. D. W. Ewick, "Finite Difference Modeling of 
Laser Diode Ignited Components", Proceedings 
of the Fifteenth International Pyrotechnics 
Seminar, Boulder, CO, 9-13 July 1990, p. 277. 



- 69 - 






Laser Diode Ignition (LDI) 

William J. Kass, Larry A. Andrews, Craig M. Boney, Weng W. Chow, 

James W. Clements, John A. Merson, F. Jim Salas, Randy J. Williams 

Sandia National Laboratories 

Albuquerque, NM 

and 

Lane R. Hinkle 

Martin Marietta Speciality Components 

Clearwater, FL 



/ 



i 



ABSTRACT 
This paper reviews the status of the Laser Diode Ignition (LDI) program at Sandia National Labs. 
One watt laser diodes have been characterized for use with a single explosive actuator. Extensive 
measurements of the effect of electrostatic discharge (ESD) pulses on the laser diode optical 
output have been made. Characterization of optical fiber and connectors over temperature has 
been done. Multiple laser diodes have been packaged to ignite multiple explosive devices and an 
eight element laser diode array has been recently tested by igniting eight explosive devices at 
predetermined 100 ms intervals. A video tape of these tests will be shown. 



INTRODUCTION. 
Laser diode ignition of explosive 
ordnance[l],[2] is an active program at 
Sandia [3]. Optical ignited ordnance 
enhances safety both on a component and 
system level. Electrically initiated devices 
can be sensitive to electrostatic discharges 
which dictate special handling care. 
Accidental initiation could cause personal 
injury or death. Optical ignition eliminates 
the possibility of this occurrence. Other 
advantages of optical ignition are resistance 
to triggering by electromagnetic radiation, no 
electrical conductance after fire and the 
absence of corrodible electrodes or 
bridgewires. 

The goal of this program is to develop a 
laser diode based optical firing system to 
ignite an octahydro- 1,3,5, 7-tetranitro- 
1,3,5,7-tetrazocine (HMX) /carbon mixture 
in an explosive actuator. Several 
components and systems have been built and 
tested. Among these are a single laser diode 
system, a three laser diode system, a high 
power laser diode igniting several actuators 



"simultaneously", and an addressable array of 
laser diodes used to ignite multiple 
detonators. The building blocks of these 
systems, environmental testing of the 
components and the various systems and 
their function will be described. 

SINGLE LASER DIODE FIRING 

SYSTEM. 

The simplest system which we have 

developed consists of a single laser diode, an 

optical fiber, a connector and an explosive 



Laser 
Diode 



f X14.5TC 
Las 
X2.0 T-t- 



Optical 
Fiber 

Connected 



Optical 



Optical Power 

TOTAL 11.8d B 

Laser Degradation 

" ' Environmental3.0dB 



UD1 
Fib 



er 



X2.0 Coupling Loss 3 QdB 

X1.01FiberLoss 01dB 

X1.2 Connection Loss 0.8dB 
X1.01FiberLoss 0.1dB 



Explosive ■ X3.0 Fire Reliability 4.8dB 
Figure 1. Power budget for laser diode ignition. 



- 71 - 



PAGE. 



± 



actuator. This system along with a power 
budget for each loss element is shown 
schematically in Figure 1. 

The losses indicated in Figure 1 are 

referenced to the nominal (50%) fire level of 
the explosive actuator. A factor of 3 .0 
(4.8dB) is arbitrarily used to achieve a higher 
fire reliability. The actual fire reliability will 
require a measure of the spread in the 
measurement of the fire threshold. Fiber 
losses (0.1 dB) and connector losses (0.8dB) 
are estimated from our experience with 
commercial connectors and fibers. The 
coupling loss (3.0 dB) represents the worst 
case coupling between the laser diode chip 
and the integrated optical fiber with which it 
is packaged. Degradation over time, 
temperature and thermal and mechanical 
environments is taken as a factor of two (3.0 
dB) over the expected lifetime (20+ years) of 
the laser diode ignition system. The result of 
this analysis is that the laser diode chip 
power, before coupling, necessary to ignite 
the explosive is approximately 15 times the 
nominal fire threshold. For HMX/C, the 
nominal threshold for a 10ms optical pulse is 
70 mW, hence, a 1.0+ watt uncoupled laser 
diode chip is required. Commercially 
available laser diode chips delivering greater 
than LOW coupled power have been 



LMHMtdfl RAIM; Wrt »*p 1T M 




Figure 2 Hermetically sealed-fiber optic 
coupled 1W laser diode for optical ignition. 















^** 














,.->--* 
















S^" 












^^> 


^>C" ^" 










,y<<,' 














1&s* 


!5?^x 












-£%& 


S^^^"" 












d 
















&s& 














^& 


Vfrl 













CURKIHT (A) 



Figure 3. Optical power vs. drive current 
at tempertures between -55C and 75C for 
a typical 1W laser diode. 

obtained and evaluated. 

The single laser diode firing system also 
includes an electronic drive circuit used to 
convert 28 V-10 ms input power pulse to the 
1.6-2.0 A- 10 ms drive current pulse required 
to deliver 1 W optical power from the fiber 
coupled laser diode. 

LASER DIODE. 

The laser diodes used for these tests have an 
optical output of approximately 1 W in the 
spectral range of 800-850 nm. The laser 
diode is an AlGaAs single quantum well 
device manufactured by Spectra Diode Labs. 
The laser diode is hermetically packaged 
with an integrally coupled 0.22 numerical 
aperture- 100 |im diameter optical fiber 
which is terminated in a commercial optical 
connector ferrule. A photo of this device is 
shown in Figure 2. 




Figure 4 Spectrum for a typical AlGaAs 
quantum well laser diode. 



- 72 - 



1 T 

0.9 
0.8 
0.7 

\ 0.8 • ; 
! 0.5 • • 
JO.*- 

0.3 

0.2 

0.1 ■: 
- 



Y/vWvW 




10 15 

Temperature Sequence 



Figure 5 Optical transmission during 
temperature cycling for an ST type connector. 

Figure 3 shows the optical power vs. drive 
current for a typical 1 W laser diode used for 
this system. The coupling efficiency from 
the chip to the fiber is greater than 50% and 
the nominal power at room temperature is 
about 0.8 W at 1.5 A drive current. At 75°C, 
however, the power has been degraded to 
0.6 W, still in excess of the power required 
from the power budget. 

Because of the broad band absorptance of 
the HMX/carbon [4] mixture used for LDI, 
the output spectrum of the laser diode is less 
important than the output power. However, 
the spectrum is an indicator of the proper 
function of the diode [5], Figure 4 shows 
the spectrum for a typical diode laser. 

OPTICAL FIBER AND CONNECTORS. 
The diode package fiber is connected to the 
explosive actuator via commercial optical 
connectors and fiber. The optical connectors 
have been tested over temperature by cycling 
between -55°C and 100°C multiple times. 
As can be seen in Figure 5, the transmittance 
degrades through the first two cycles and 
then oscillates between high and low 
temperature values. The optical fiber used 
was 0.22 NA, 100 |am diameter, pure silica 
core, doped silica cladding obtained from 
Polymicro Technologies. The results of 
temperature testing only this fiber are shown 




Figure 6 Optical transmission vs. 
temperature for Polymicro optical fiber. 

in Figure 6. It can be seen that a reversible 

transmission loss occurs as the fiber is cycled 
to low temperature. The loss is consistent 
with losses predicted from microbending due 
to the mismatch in thermal expansion 
between the fiber and the polyimide coating 
[6]. A larger diameter Polymicro fiber and a 
fiber from General Optics (with a loose 
acrylate buffer) were also tested as shown in 
Figure 7. The losses from the 400 \im 
Polymicro fiber were less pronounced than 
the 100 \im while the losses from the 
General Fiber were negligible. 

EXPLOSIVE ACTUATOR. 
The majority of the explosive 



1 T 

Ofl ■ 

ot ■■ 

j 0.6 
! 0.4 
' 0J- 

02 

0.1 




Genere) Rber Optic* 100/140 Ffc*r **i Aciyfate 




Pdyrrtcre 40W440 fiber wHh PdyknMa Buflbr 



I Rbwt ar«32 FmI Long In 10 bich Dtwnttor coil 



-20 -10 



Figure 7 Optical transmittance vs. temperature 
for two kinds of optical fiber. 



- 73 - 



characterization done at Sandia National 
Labs has been for 2-(5 cyanotetrazolato) 
pentaamine cobolt III perchlorate (CP) and 
TiH 165 KC10 4 ignition charges. TheCP 
ignition charges are normally the first 
element in a detonation column consisting of 
1.7 g/cm 3 CP doped with carbon black, 
1.5 g/cm 3 CP, 1.7 g/cm 3 HMX. The charges 
are 20 mg of material in a 2. 1 mm diameter 
by 2.5 mm long cylinder. They are unsealed 
and the fiber is placed in direct contact with 
the explosive powder. The LDI system was 
designed to operate with explosive actuators 
to generate gas and perform mechanical 
work. The actuators consist entirely of a 
mixture of HMX and 3% carbon black. The 
carbon increases the material absorptance in 
the near IR where the laser diode emits. 

ESD TESTING 

Laser diode ignition derives its immunity to 
ESD and electromagnetic radiation (EMR) 
because of the absence of electrical 
conductors within the region where the 
energetic material is located. A related issue 
in ESD safety, however, is what the optical 
output of the laser diode is when it is 
subjected to an ESD pulse. Is the optical 
energy or power generated sufficient to 
ignite the energetic material? To address this 
issue we measured the optical output while 
subjecting the laser diode to an ESD pulse. 
This pulse is defined by an electrical circuit 



25 

20 

s« 






SOL RK347 


































£ 10 
5 






















S 10 15 20 t 

Ttfmfrtt) 



Figure 9 Optical output power for a laser 
diode subjected to a series of Sandia Severe 
Human Body ESD pulses. 



designed to simulate a human body spark 
including a hand (small capacitance) and 
body discharge [7]. The schematic of the 
measuring equipment is shown in Figure 8 . 

A series of measurements were made by 
increasing the voltage on the SSET (Sandia 
Severe ESD Tester) circuit and monitoring 
the current through the laser diode. The 
optical output from the laser diode was 
coupled to an optical fiber and measured 
with a fast response photodetector. The 
early time results are shown in Figure 9. 
The peak output is reached in 2 ns followed 
by signal decay with two time constancts 
The first decay constant is a few 
nanoseconds and the second is 300 ns. 
These decay constants result in a total signal 



HIGH 
jVOLTAG 
SUPPLY 



Scope] 

i> .W.".U. ' .". ' , I M I * 













Figure 8 Test setup for measuring laser 
diode output power with an ESD pulse 
current input. 











































A 




















\\ 




















■\» 


<r 


















V 


X 


















. 10KV 


^^ 


^ 

























































200 H0 400 



Figure 10 Long time decay of optical output 
for 25kV, 15.5kV and lOkV circuit input 
voltages. 



- 74 - 



30 








SDL RK347 








150 


25 


















125 


5-2C 








- - — 


_~-- 


- ~2^* 






140 


in 



















4 

o 

50 


o '* 

a. 
1( 


















5 


















25 

























10 



12 



14 



10 18 20 22 
Charge Voltage (kV) 



24 



26 



Figure 11 Peak output power (broken line) 
and current (solid line) vs. charge voltage on 
ESD test circuit. 



decay in about 1000 ns. As the input voltage 
to the ESD circuit is increased from 10 kV 
to 25 kV, the peak current increases from 
60 A to 140 A. The optical output tracks 
this current until degradation of the diode 
begins. This occurs at about 125 A with a 
circuit input voltage of 23 kV. The peak 
optical output power is 25 W while the 



1 fZ.-r\j\j - 


i 










1 fc-v 1 

1 c_no . 












1 c.-\jz • 








yj \r 


1 C./T4 . 








^ 


P 


i t-u j ■ r 




J 


v 




^ 1 F J14 . 






/ 






>» 

c 1 F_fi«s . 


! 










1 F-ftR . 


[ 










1 t*vw ' 
1 C_fl7 . 




ESD 








1 fc-U/ * 












1 P-Tlft - 












1E-09 - 











1E-09 1E-07 1E-05 1E-03 1E-01 1E+01 
Time (s) 

Figure 13 Comparison of optical energy 
necessary to ignite TKP and CP with the 
maximum optical energy available from an 
ESD source. 



1E+02 r 



1E+01 



1E+00 



8 
O 



1E-01 



1E-02 



: 


















"> 


\ 












:bSL 


"^ 


\ 






V 














^ 


^ 


*XKF 














I 


S.«r*i> 
















\OI 



1E- 1E- 1E- 1E- 1E- 1E- 1E- 1E- 1E- 
09 08 07 06 05 04 03 02 01 

Time (s) 

Figure 12 Optical power necessary for 
explosive ignition and optical power 
generated by ESD pulses vs time. 



maximum energy found from integration of 
the complete decay is 5 ^J. The optical 
decay until 1000 ns is shown in Figure 10. 



Figure 11 shows peak optical power and 
diode current plotted vs. circuit input 
voltage. The behavior of this laser diode and 
others tested indicates that even though the 
current through the diode continues to 
increase with higher circuit drive voltages, 
the optical power levels off and begins to 
decay as the laser diode is degraded. This 
phenomenon becomes a safety advantage 
because the diode will not be able to deliver 
sufficient power to ignite the explosive. 

Figure 12 shows the optical power available 
from a laser diode driven by an ESD source 



- 75 - 



been built and tested in various 
configurations. We have estimated a power 
budget for reliable ignition and the individual 
loss terms are being characterized. Fiber 
losses and connector losses can be 
accommodated with proper choices of 
connectors and fibers. Currently available 
commerical high power laser diodes provide 
ample power to ignite a variety of explosives 
under extremes in temperature and 
mechanical environments. The behavior of 
these laser diodes is being characterized over 
temperature, mechanical environments and 
time (aging). ESD testing demonstrates that 
the laser diode is inherently safe from 
producing optical power or energy which 
exceeds the explosive ignition threshold. 

ACKNOWLEDGMENT 

The authors would like to express their 
appreciation to John Barnum for making the 
ESD measurements. 



vs. time and compared to the power-time 
combination required to ignite titanium 
potassium perchlorate or CP. The HMX 
threshold falls slightly below that for CP. It 
can be seen from this plot that even though 
there is ample power for ignition, the 
duration is too short to ignite the explosive 
material. 

The integration of the power vs. time curves 
gives a maximum available optical energy of 
3-5 ^J from this type of ESD pulse. This 
energy also probably represents the 
maximum energy available under other high 
current conditions. Figure 13 shows the 
available optical energy plotted vs. time 
compared to the energy required to ignite CP 
or TKP. When plotted in this manner it is 
apparent that too little ESD generated 
optical energy is available to ignite these 
explosives. 

CONCLUSION 

A complete laser diode ignition system has 

REFERENCES 

1. S. C. Kunz and F. J. Salas, "Diode Laser Ignition of High Explosives and Pyrotechnics, Proc. 
of 13th International Pyrotechnics Seminar, Grand Junction, CO, 1 1-15 July 1988, pp505 

2. D. W. Ewick, L. R. Dosser, S. R. McComb and L. P. Brodsky, "Feasibility of a Laser Ignited 
Pyrotechnic Device", Proc of the 13th International Pyrotechnic Seminar, Grand Junction CO, 1 1- 
15 July 1988, pp 263 

3. J. A. Merson, F. J. Salas, W. W. Chow, J. W. Clements and W. J. Kass, "Laser Diode Ignition 
Activities at Sandia National Laboratories", Proceedings of the First NASA Aerospace 
Pyrotechnic Systems Workshop, Houston, TX, 9-10 June 1992, ppl79-196. 

4. R. J. Jungst, F. J. Salas, R. D. Watkins and T. L. Kovacic, "Development of Diode Laser- 
Ignited Pyrotechnic and Explosive Components," Proc. of the 15th International Pyrotechnics 
Seminar, Boulder, CO, 9-13 July 1990, pp549-568. 

5. L. F. DeChiaro, S. Ovadia, L. M. Schiavone, C. J. SandrofF, "Quantitative spectral analysis in 
semiconductor laser reliability," SPIE Technical Conference 2148A, Los Angeles, CA, 24-26 
January, 1994. 



- 76 - 



6. Powers Garmon, "Analysis of Excess Attenuation in Optical Fibers Subjected to Low 
Temperatures", Proc. of the International Wire and Cable Symposium, 1983, ppl34-143. 

7. R. J. Fisher, "The Electrostatic Discharge Threat Environment Data Base and Recommended 
Baseline Stockpile-to-Target Sequence Specifications," SAND88-2658, November 1988. 



- 77 - 



STANDARDIZED 
LASER INITIATED ORDNANCE 






James V. Gageby l 

Engineering Specialist 

Explosive Ordnance Office 

The Aerospace Corporation 

Abstract 

Launch vehicles and spacecraft use explosively initiated devices to effect 
numerous events from lift-off to orbit These explosive devices are electrically initiated by 
way of electro-mechanical switching networks. Today's technology indicates that 
upgrading to solid state control circuits and laser initiated explosive devices can improve 
performance, streamline operations and reduce costs. This paper describes a plan to 
show that these technology advancements are viable for Air Force Space and Missile 
System Center (SMC) program use, as well as others. 



Introduction 

A plan to develop, qualify and flight 
demonstrate a laser initiated ordnance system 
(LIOS) has been accepted by the SMC Chief 
Engineer and The Aerospace Corp. Corporate 
Chief Engineer as part of their horizontal 
engineering program. The Chief Engineers' 
horizontal engineering effort includes a task for 
standardization of systems and components 
common to a variety of programs. The objective of 
standardization is to reduce costs by eliminating 
duplications in development and qualification often 
seen when vertical engineering prevents cross 
pollination. 

The LIOS is intended as a state-of-the-art 
solid state replacement for the present day 
electrically initiated ordnance firing circuits for 
future space launch vehicle and satellite systems. 
The LIOS eliminates the need for electro- 
mechanical safe and arm devices and latching 
relays that are presently used in today's ordnance 
firing circuits. 

This plan will result in confirmation of LIOS 
suitability for SMC applications. It will establish a 
performance and requirements specification for 
standardization on SMC programs. Flight system 
performance enhancements and cost savings will 
result from the safety improvements, streamlined 
operational flow, weight savings, improved 
reliability and hardware interchangibility features of 
this new technology. It is expected that a family of 
LIOS's having various multiple output 
configurations will be developed to fit SMC 
program needs. 

A typical SMC launch vehicle and satellite 
uses at least 40 explosively initiated events to get 
into proper orbit. The majority of these are 
redundant, therefore, 80 explosive initiations can 



occur from engine ignition and lift-off to final 
appendage deployments in orbit. At the extreme 
NASA's space shuttle uses more than 400 
explosive events from lift-off through deployment 
and release of their drag parachute on landing. 

Shown below is a simplified description of 
the conventional ordnance firing circuits used on 
most SMC programs to effect these explosive 
initiations. 

Conventional Ordnance Firing Circuit 



Electro-mechanical device with EED 
and explosive train Interrupt (moving 
parts) - Range Safety requirement for 
FTS and SRM ignition 



Power/Control 
Input 




Cu Wire 



Electro Explosive Device (EED) (no moving parts)-* 
Range Safety assessment needed for potentially 
hazardous applications 

In the conventional system, sequenced 
power and control inputs from system computers 
are routed to a switching network that allows safe, 
arm and fire commands to be sent to the explosive 
devices. Mechanical latching relays are used to 
effect these commands. 

The commands are sent via copper wire to 
either an electro explosive device (EED) or to a 
safe and arm device that contains an EED. The 
EED has an electrically conductive path directly to 
the explosive materials internal to it. Electrical 
energy in this path, at predetermined thresholds, 
causes EED ignition. The EED contains no moving 
parts. 

The safe and arm device (S/A) is an 
electro mechanical component required for 



- 79 - 



.fttSc. 



1« 



compliance with safety regulations in flight 
termination and solid rocket motor ignition systems 
only. It contains moving parts. It provides a barrier, 
or interrupt, in the explosive train so that premature 
ignition of the EED will not cause an unplanned 
event. This interrupt is remotely removed during 
the mission sequence to allow end item function. 
The safe and arm device also contains a 
component called a safing pin which must be 
manually removed before remote arming and firing 
can be effected. Removal of the safing pin is done 
late in the pre-launch cycle and usually requires 
the launch site to be cleared of all but essential 
personnel. 

The LIOS replaces the EED used in the 
conventional ordnance system with a device that 
uses laser diode energy to ignite the same 
explosives. The primary advantage is the 
elimination of electrically conductive paths to the 
explosive mixes. This drastically reduces concerns 
of premature ignition since external environments 
like static electricity, electro-magnetic interferences 
as well as radio frequency (R-F) fields are isolated 
from the explosives. 

The new explosive component is called a 
laser initiated device or LID. The LID will be 
designed to use secondary explosive materials as 
ignition sources rather than primaries as used in 
EED's. This reduces handling concerns. The LID 
could be considered in the same category as small 
arms ammunition for handling and shipping 
purposes. This will result in a significant, although 
indeterminate, cost savings. 

The LID outputs can be configured to be 
nearly identical to the EED outputs; therefore, 
interfaces with present day explosively actuated 
components will be compatible. Requaiification of 
explosively actuated components with LID's, that 
were previously qualified with EED's, should be 
minimal. 

LIOS Concept 

A description of the LIOS is given below. 



LFU (LASER GENERATOR) 



ELECT POWER t 
SIGNAL CONTROL I 



LIOS 

_- e LID (EXPLosrvE) 

ETS (FIBER OPTIC PATH) l 



Z^-\ 



Laser firing unit (LFU) - receives control signals and power 
(28VDC) for sequencing. Laser diodes in LFU produce single 
or multiple laser outputs. Contains no moving parts. 

Energy transfer system (ETS) - conveys laser light through 
fiber optics and connectors to LID. 

Laser initiated device (LID) - allows laser energy (heat) to be 
absorbed into chemical mixture causing deflagration/ 
detonation of explosive. 



The LIOS contains no moving parts. The 
switching network in the LIOS uses solid state 
electronics to accomplish the functions mechanical 
latching relays and S/A's provide in the 
conventional ordnance system. 

Advantages of LIOS are in reduced 
handling and safety concerns during ground 
operations. During most pre-launch operational 
cycles, R-F silence and limited access conditions 
are in effect while ordnance installations are in 
progress. These down times could be eliminated or 
drastically reduced by use of the proposed LIOS. 
The reduced safety concerns may allow for 
installation of ordnance items at the factory instead 
of the launch site, thus reducing pre-launch 
operational costs by streamlining ground 
operations. 

The use of lasers for ignition of explosives 
is not new. Research in this area began more than 
twenty-five years ago. In fact, the small 
intercontinental ballistic missile (SICBM) program 
developed and flight demonstrated a crystalline rod 
laser system a few years ago. The SICBM laser 
ordnance concept is shown below. 

SICBM Laser Ordnance Concept 

Fiber 

V OP**"/ LID 



For clarity 

features are not shown 

(moving parts required) 

Rod Laser 



Motor Driven Sequencer 




Safe/Arm with optical shutter 



The SICBM concept served it's purpose 
well. Unfortunately, it is more complicated than the 
conventional ordnance firing system. It requires 
numerous electro-mechanical components for 
safety and operational reasons. 

Power to the rod laser is sequenced 
through an electro-mechanical switching network 
similar to that used in the conventional system. For 
system function an optical shutter is remotely 
actuated to allow lased light to be directed onto a 
motor driven sequencer. The sequencer has 
optical prisms on a rotating wheel that allow for 
splitting of the laser beam for multiple output 
functions. 

Included in the system, but not shown in 
the schematic, is a built-in-test (BIT) feature. The 
BIT feature requires additional electro-mechanical 
devices to bypass the rod laser and let a light 
emitting diode (LED) be pulsed into the fiber optic 
transmission line. To verify health of the 
transmission line, the LED's pulse is reflected at 



- 80 - 



the LID and it's total travel time measured by 
means of an optical time domain reflectometer 
(OTDR). This transmission line check-out is done 
as late in the pre-launch cycle as practical. The 
OTDR is not part of the flight hardware. 

The use of semiconductor laser diodes as 
an ignition source is a new, emerging technology. 
Their use is, by all indications, a viable alternative 
not only to the present day electrically initiated 
systems but also to the SICBM approach. At the 
onset of the SICBM effort, laser diode technology 
had not developed sufficiently to provide output 
energies needed to meet LID ignition margin 
requirements. Today the technology has 
progressed to a point that laser diodes can provide 
high energies with ample margin. 

Using laser diodes in place of crystalline 
rod lasers is a quantum leap in miniaturization. 
This miniaturization allows for multiple outputs 
without having to use a mechanical prism 
sequencer as in the SICBM system. All mechanical 
components are removed. It also allows for the 
introduction of solid state control logic circuits to 
further advance explosive ignition technology in 
space applications. 

SMC LIPS 

The SMC LIOS standardization plan is 
broken down into six major tasks shown 
below. 

LIOS Major Tasks 



Task 



1. Acquire Range Cmdrs 
Council approval for 
use of solid state 
UOS 

2. System and circuit 
modeling 



3. Determine system 
compatibility with 
RF/EMI/ESD 

4. Verify compatibility 
with SMC programs 

5. Determine cost 
benefit of LIOS use 

6. Qualify and flight 
demo an SMC 
compatible LIOS 



Accomplishment 



Eliminate mechanical 
components In 
ordnance firing circuits 

Validate circuit 
performance 

Validate system 
energy margins 

Assess BIT designs 
Compatibility validated 



Interfaces validated 



Cost benefits defined 



LIOS technology ready 
for SMC use 



Exit Criteria 



LIOS not approved or use 
of moving components 
with LIOS mandated 

Solid state logic cannot 
meet performance/ 
safety requirements 
Lack of margin 

BIT not compatible 
with designs 

System not compatible 



UOS not compatible 

No cost savings 
Funding not available 



The following discussion will outline the 
key points of each task. Note that the exit criteria 
shown for each task is not task completion. It is 
criteria that will prevent LIOS from becoming a 
standard for SMC programs, i.e., criteria that would 
cause cessation of the SMC LIOS standardization 
effort. 



The first Task is to obtain an agreement 
with the Range Commanders Council allowing use 
of LIOS at all launch sites. To be specific, the 
agreement must allow use of a LIOS, without 
moving parts, i.e., remotely controlled shutters, 
etc., on any ordnance system, at any launch site. 
This includes both flight termination and 
operational ordnance firing systems. 

If the Task 1 agreement cannot be attained 
the SMC LIOS effort will stop. The cost advantages 
of a LIOS using mechanical components compared 
to the cost of today's conventional ordnance firing 
system are not of sufficient magnitude to warrant 
implementation. 

The remaining tasks will provide technical 
rational to support safety and performance 
requirements of the Task 1 agreement. These must 
satisfy any Range Commanders Council or SMC 
program concerns. 

The second task is to analyze the solid 
state circuits to verify that they can meet safety and 
performance requirements. This will be followed by 
an effort to model the entire LIOS and assess 
performance margins. The margin analysis must 
show that there is at least 50% more energy 
available than necessary to ignite the LID when all 
system parameters and external environments are 
at their extremes. If the designs can not show 
sufficient margins for safety and performance 
needs, the SMC LIOS effort will be stopped. 

Circuit concepts will be analyzed and be 
validated by bench tests of designs that are 
representative of the optimum configurations. 
These tests are considered a key element in the 
validation of the LIOS concept. Contributing 
expertise to these tasks are Dave Landis and Don 
Herbert of the Electronics Division. 

During the course of the modeling BIT 
designs will be evaluated. The BIT feature will be 
used to check continuity of the ETS path between 
the laser diode and the LID only. Full power system 
checks will be done prior to final connection of the 
LID and be performed as late in the pre-launch 
cycle as deemed practical. A key feature of the 
LIOS implementation is an ability to perform 
remote check-out of system health without 
interfering with other pre-launch activities. 
Therefore, the LIOS effort will be stopped if an 
adequate BIT feature cannot be found. 

The third task is to determine the LIOS 
compatibility with external environments that may 
cause premature ignition or prevent ignition of the 
LID. These environments include lightning induced 
electro-static discharges (ESD), R-F and electro- 
magnetic interferences (EMI). These are the same 



- 81 - 



environments that are concerns for conventional 
ordnance firing circuit designs. As previously 
noted, current designs are influenced by these 
while the LIOS is not. Verification of LID 
compatibility with reasonable limits of these 
environments is obtainable. Much work has been 
done in this area and will be evaluated for 
applicability. 

The LFU must also be shown to 
adequately shield these environments from the 
sensitive components within it. If the LID or the 
LFU can not be shown to survive reasonable limits 
of these environments, and designs cannot be 
altered to do so, the SMC LIOS effort will be 
stopped. 

The fourth task examines the compatibility 
of LIOS with common SMC program interfaces. An 
attempt will be made to determine the optimum 
LIOS configuration in terms of the number of LID 
outputs, control circuit configurations and BIT 
options. This will, more than likely, result in several 
configurations and create a family of LIOS options. 
Determining the number of changes to the LIOS 
design to suit interface needs and maximize 
standardization will be a major part of the task. 

All of the above will have a direct impact 
on Task 5 which will evaluate cost benefits of LIOS 
implementation. Task 5 is also affected by other 
factors including ground operations and flight 
performance improvements. In ground operations 
costs, procedural changes in handling and check 
out of conventional ordnance systems versus LIOS 
need to be assessed. 

It is anticipated that use of LIOS on SMC 
programs will be limited to new programs and to 
those undergoing major changes. The non 
recurring costs of a blanket change to use a LIOS 
on existing programs is prohibitive. No other 
justification would out weigh these cost differences. 

If the fifth task indicates that there is no 
cost savings the effort will be stopped. Likewise, if 
Task 4 shows that the LIOS is not compatible with 
SMC programs the LIOS effort will be stopped. 

The sixth task will be the ultimate proof of 
the LIOS concept and it's compatibility with SMC 
programs. The work of the other tasks will result in 
creation of a performance and requirements 
document that will be used to solicit multiple 
suppliers for qualification of LIOS designs. This will 
be followed by a flight demonstration on an SMC 
program. Success will demonstrate the usefulness 
of LIOS for space and launch applications. Task 6 
will not be executed if funding is not made 
available. 



Acknowledgments 

The author wishes to thank Col. J. 
Randmaa, Col W. Riles, Maj. K. Johnson, Capt. R. 
Anderson and W. Evans of the SMC Chief 
Engineers office, and Dr. J Meltzer, Dr. R. Hall and 
J. Gower of The Aerospace Corp. Chief Engineers 
office for sponsoring the pursuit of the LIOS plan. I 
also need to thank Norm Schulze of NASA Hqtrs 
for the opportunity of being a member of his 
NASA/DOD/DOE Pyrotechnic Steering Committee 
where the LIOS concept was initiated. 



- 82 - 



1-2% 



~1 



v/1 



MINIATURE LASER IGNITED BELLOWS MOTOR 

Steven L. Renfro 

The Ensign-Bickford Company 

Simsbury, CT 

Tom M. Beckman 

The Ensign-Bickford Company 

Simsbury, CT 



Abstract 

A miniature optically ignited actuation device has 
been demonstrated using a laser diode as an 
ignition source. This pyrotechnic driven motor 
provides between 4 and 6 lbs of linear force 
across a 0.090 inch diameter surface. The 
physical envelope of the device is 1/2 inch long 
and 1/8 inch diameter. This unique application of 
optical energy can be used as a mechanical link 
in optical arming systems or other applications 
where low shock actuation is desired and space 
is limited. 

An analysis was performed to determine 
pyrotechnic materials suitable to actuate a 
bellows device constructed of aluminum or 
stainless steel. The aluminum bellows was 
chosen for further development and several 
candidate pyrotechnics were evaluated. The 
velocity profile and delivered force were quantified 
using an non-intrusive optical motion sensor. 



Intrpduptipn 

A small optical to mechanical link has 
been developed for uses where low 
velocity force is required. This device 
uses a small B/KN0 3 charge to actuate 
a miniature rolling bellows. This laser 
diode ignited system provides 
approximately 4 lbs of force over 0.1 
inches of displacement. The device was 
designed to move a small barrier either 
into or out of the way to provide means 
for a miniature optical arming feature. 
The overall size of the device is less 
than 1/2 inches long and 1/8 inches in 



diameter. A mounting feature allows 
simplified integration into new or existing 
systems. 

Design Analysis 

The challenges of this development 
program are to balance the gas output to 
desired force, ignite with a 1 Watt rated 
laser diode, and to downsize processes 
to manufacture such a small complex 
device. 

The force is dictated by the material used 
for the bellows. This analysis assumes 
that in order to sufficiently move the 
bellows, plastic deformation must occur. 
This requires that the yield point of the 
selected material be exceeded without 
violating the ultimate strength. Following 
these guidelines, the pressure required 
for actuation can be calculated. The 
results are summarized in Table 1. 

The burst pressure for hardened 
aluminum alloys is less than the pressure 
required for actuation. Annealed 302 
Stainless Steel, Aluminum 1100-0, or 
Aluminum 3003-0 are suitable bellows 
candidate materials. Using this 

information, the resultant force developed 
by a fully actuated bellows can be 
calculated. The force results are listed in 
Table 2. 

In order to produce the desired force, 



- 83 - 



several candidate pyrotechnic materials 
and stoichiometrics were considered. 
Size restraints required that the selected 
pyrotechnic use the space allocated for 
the charge holder precisely. This 
analysis was crucial due to space 
restraints for the charge allocated given 
the overall size envelope. 

The amount and type of pyrotechnic 
material was calculated based on the 
pressure required for actuation using the 
NASA-Lewis equilibrium thermochemistry 
code. The results of this analysis are 
normalized to Ti/KCI0 4 and are listed in 
Table 3. 

The mass calculations were used to 
select materials for prototype testing. 
Based on these mass calculations, the 
first candidates selected for prototype 
testing were B/KN0 3) Ti/KCI0 4 , and 
B/BaCr0 4 /KCI0 4 . 

Initial Prototype Testing 

In order to gain information isolated to 
function of the bellows, larger prototypes 
were used for the initial test series. The 
results of the first two groups of five 
prototypes each are listed in Table 4. 

The B/KN0 3 resulted in an acceptable 
charge weight for the desired extension. 
The other candidates did not perform 
successfully. The B/BaCr0 4 /KCI0 4 
loaded devices would not ignite using 
the output from a 1 W laser diode. The 
Ti/KCL0 4 loaded devices resulted in 
burst of the bellows. The burn rate of 
the Ti/KCL0 4 did not provide the low 
velocity required for this application. 

Sensitivity of B/KNO, 

The B/KN0 3 material ignites consistently 



using full power 840 nm diode with a 10 
ms pulse width. An interface sensitivity 
test was used to verify reliability. The 
results of this testing using 200 micron 
fiber are listed in Table 5. These results 
are listed based on the calibrated output 
from the diode and do not include line 
losses. 

Process Development 

The success of this miniature component 
depends highly on an integral charge 
holder / fiber optic subassembly. In 
order to offset losses expected in fiber 
optic interfaces, a smaller core fiber was 
chosen to increase the power density of 
the available optical energy. This allows 
for the fiber to be prepared prior to final 
assembly into the charge holder. 
Polishing would not be possible given the 
restricted space. The Ensign-Bickford 
Company developed a cleaving 
technique capable of limiting losses to 
less than 1dB at the final assembly level. 
These losses are acceptible for reliable 
ignition without polishing the fiber in the 
final assembly. 

The charge for the test units was 
pressed directly onto the fiber to ensure 
intimate contact between the fiber and 
the B/KNO3. 

The development units were assembled 
using processes developed for 
miniaturization. These early development 
units were then functionally tested to 
verify analysis and prototype work and to 
determine force and velocity output. 

Force Output Testing 

The units are designed to function under 
axial load, therefore, is is desirable to test 
them in that mode. A crushable foam 



- M - 



was selected to determine the 
approximate output force developed by 
each test unit. The bellows must develop 
at least 2.64 lbs to actuate. The goal for 
total nominal developed force is 4.02 
lbs. A polyurethane foam was selected 
with a minimum compressive strength to 
require at least 2.0 lbs to crush in order 
to assess the total nominal force output. 
Figure 2 illustrates the test setup and 
Figure 3 illustrates the results of the four 
units tested using this method. 

Each of the test articles were bonded to 
the test block utilizing the existing 
mounting flange and two part epoxy. 
This bonding method successfully held 
each test unit during function. Three of 
the units extended approximately 0.060 
inches into the foam block and the fourth 
did not actuate. The failed unit was 
inspected and revealed that the bellows 
had been inadvertently bonded in place 
during assembly. This unit burst under 
the developed pressure. This lesson 
learned resulted in careful inspection of 
the bellows area after final assembly. 
The area filled with epoxy cannot be 
readily viewed with the unaided eye. 
Future assembly will necessarily require 
magnification. 

Velocity Testing 

To assess the overall impulse of the 
delivered force, a simple velocity 
measurement sensor was devised. This 
article is illustrated in Figure 2. 

The velocity fixture is quite simple. Two 
donor fiber optics are aligned across a 
channel with acceptor fiber optics. Each 
donor / acceptor pair is placed at a 
known distance from the unextended 
bellows. Using white light as a source, 
the acceptor fiber picks up the light and 



is then connected to a photo-diode to 
produce a small voltage. Upon function 
of the unit, this optical path is broken 
resulting in a voltage drop across the 
photo-diode. This voltage is monitored 
using an oscilloscope to determine a time 
difference between the donor / acceptor 
pairs. This measurement scheme allows 
for determination of average bellows 
velocity without interrupting the function. 

The results of three of these test units 
are presented in Figure 4. During 
function of the bellows motor into air, the 
test bellows for the first unit did not stay 
intact. This indicated that the unit 
probably is producing too much gas for 
function without axial load. The resultant 
velocity is not for the entire bellows for 
this test unit, but for the aluminum end 
free from the assembly. The second and 
third units functioned correctly and the 
velocities measured are for the bellows. 

Further Development Work 

The Ensign-Bickford Company is 
continuing to develop this product under 
contract for Los Alamos National 
Laboratory. The ideal miniature bellows 
will function under load to produce the 
desired force and be able to function in 
air without expelling products of reaction. 
The final development phase is to 
concentrate on optimizing the charge 
size in order to meet these goals. 

Discussion 

The analysis and prototype phase 
contributed to development of the 
miniature bellows motor. More work 
needs to be done to refine the design. A 
pyrotechnic device to deliver a small 
amount of force is possible and has been 
demonstrated. 



- 85 - 



Table 1. Pressure Requirements for Bellows Actuation and Burst 



Bellows 
Material 


Alloy and Temper 


Minimum 

Actuation 

Pressure 

(psi) 


Burst 

Pressure 

(psi) 


Aluminum 


1100-0 


389 


577 


Aluminum 


1100-H12 


1089 


689 


Aluminum 


1100-H14 


1555 


977 


Aluminum 


3003-0 


467 


711 


Stainless Steel 


302, Annealed 


2722 


4000 



Table 2. Force Developed for Various Bellows Materials 



Bellows 
Material 


Alloy and Temper 


Actuation 

Force 

(lbs) 


Target 

Force 

(lbs) 


Aluminum 


1100-0 


2.64 


4.02 


Stainless Steel 


302, Annealed 


17.32 


25.45 



Table 3. NASA Lewis Calculation Results 



Candidate Pyrotechnic 
Formulation 


Calculated Flame 

Temperature 

(K) 


Normalized Mass 

Required to 
Produce 500 psi 


Ti/KCI0 4 


5006 


1.00 


Ti/KCI0 4 + RDX 


4155 


0.63 


RDX + C 


3083 


0.67 


B/BaCr0 4 /KCI0 4 


3872 


1.19 


BKN0 3 


4044 


0.79 



- 86 



Table 4. Results of Initial Prototype Testing 



Pyrotechnic 
Material 


Charge 

Mass 

(mg) 


Ignition 
Source 


Maximum 
Extension 


BKNO3 


6 


Laser Diode 


0.08 


Ti/KCLCy 


6 


Laser Diode 


Bellows 
Burst 


B/BaCr0 4 /KCI0 4 


6 


Nd:YAG 





Table 5. Ignition Threshold Test Results Using 840 nm Diode 




Test Results 


No. of Tests 


10 


Threshold (50% Level) 


536 mW 


Standard Deviation 


56 mW 


All-Fire Level (.999/95%) 


909 mW 



- 87 - 




Figure 1 Minature Laser Ignited Bellows 



- 88 - 



. Donor Fiber Optic 



^ r^ 




1 


\ 


It 





. Acceptor Fiber Qptic 



, Foan Block 



.Bellows Extension 



Figure 2 Force and Velocity Test Setup 



- 89 - 




Figure 3 Force Output Test Results 



- 90 - 





0.012 




0.011 


^.^ 




o 




0) 

f/1 


0.01 






**■•<■•,,, 




E 


0.009 


E 




**— * 




^ 


0.008 


o 
O 

(D 




0.007 


> 




(D 
O) 
(0 


0.006 




<D 




3 


0.005 


"D 




0) 


0.004 


8 

0) 


0.003 


2 






0.002 




0.001 



Ligh t Leak 




Figure 4 Results of Velocity Tests 



- 91 - 



- 92 - 



PERFORMANCE CHARACTERISTICS OF A LASER-INITIATED 
NASA STANDARD INITIATOR 

John A. Graham 

Senior Project Engineer 

The Ensign-Bickford Company 

Aerospace and Specialty Products 



698% 



INTRODUCTION 

The Ensign-Bickford Company has been 
actively involved in the design and 
development of a laser equivalent to the 
electrically initiated NASA Standard 
Initiator (NSI). The purpose of this paper 
is to describe the present design and its 
performance characteristics. 
Recommendations for advancement of this 
program are also presented. 

DESIGN DESCRIPTION 

The Ensign-Bickford Laser-initiated NASA 
Standard Initiator (LNSI) design consists 
of an Optical Connector, Optical Fiber 
and a Propellant that is hermetically sealed 
in a Squib Housing (see Figure 1). The 
LNSI is equivalent to the NSI, using the 
NSI propellant and matching the 
installation envelope so that it can be used 
to function present NSI initiated devices. 

A standard ST or SMA connector is used 
and attached to the optical fiber with a pot 
and polish technique. The present design 
uses 200 micron Hard Clad Silica optical 
fiber but is not limited to that size; larger 
or smaller diameter fiber can be 
incorporated if dictated by system level 
requirements. The fiber is installed and 
sealed into an optical header using Ensign- 
Bickford proprietary fiber seal technology. 
The Ensign-Bickford seal has demonstrated 
hermeticity after exposure to a 40,000 psi 
proof pressure; also hermeticity is 



maintained post-function. This seal 
technology has been successfully employed 
in devices functioned from -62 °C to 
+93°C (-80"F to +200°F) and has 
successfully endured exposure to the same 
level of thermal shock without 
performance degradation. 

The propellant is the NSI-defined blend of 
Zirconium, Potassium Perchlorate, 
Graphite and Viton "B". Powdered raw 
materials are wet blended in a high shear 
blender followed by application of Viton 
"B" via a precipitation process. The mix 
has been examined by Scanning Electron 
Microscope and found to be uniform. The 
blending process does not alter particle 
morphology. 

The optical fiber is polished in situ and dB 
loss characteristics verified prior to 
propellant loading. The propellant is in 
direct contact with the exposed polished 
fiber face thus allowing laser diode power 
to reach the propellant without power 
density loss due to beam divergence. 

Hermetic sealing is provided by laser 
welds. A closure disc is laser welded onto 
the output end of the squib housing. The 
disc has a chemical milled "flower pattern" 
which "blossoms" when the device is 
functioned. The flower pattern prevents 
expulsion of large metallic particles from 
the squib and into devices which would be 
detrimental in some applications. 



PAS. 



kt; : ;?:.miy el 



- 93 - 



Two versions are available, one using a 
stainless steel housing and another using 
Inconel 718. In either design, the charge 
cavity is proof pressure tested at 15,000 
psi prior to loading propellant. 

ALL-FIRE POWER 

Reliability testing has been done to 
establish the all-fire power requirement at 
room temperature. Testing was done 
using 200jim fiber. The "pass" criteria 
was that the time from the start of the 
laser pulse to first pressure had to be less 
than or equal to 10 milliseconds, although 
the actual laser diode pulse width was 50 
milliseconds. The long duration pulse was 
used to get a better characterization of the 
relationship between power and time to 
ignition (Figure 2). The 0.9999 all-fire 
power at 95% confidence is 595 
milliwatts. This corresponds to an all-fire 
power density of 1900 watts/cm 2 . 

LASER IGNITION TRANSIENT 
THERMAL FINITE ELEMENT 
ANALYSIS (FEA) 

The reliability test data suggests that 
ignition time repeatability is a function of 
the laser power. At high laser power, the 
function time (i.e. time to ignition) is short 
and repeatable, but as power is decreased, 
function time slows and is less predictable. 

Transient FEA was done over a range of 
input powers to gain a qualitative 
understanding of this phenomena. The 
computer model included the propellant, 
fiber optic core/cladding, epoxy and 
surrounding metal structure. Upon initial 
application of laser energy, the heating 
rate is high, but slows and asymptotically 
approaches a limit that is a measure of the 
LNSFs ability to reject heat to the 
surrounding environment. If the 



autoignition temperature of the mix is 
reached quickly, then the function time 
will be very repeatable because only the 
propellant is heated and therefore only the 
variability of the propellant comes into 
play (eg mix homogeneity, density 
gradient within the pressed powder, etc). 
As the time to ignition increases, the 
thermal effects of the surrounding 
materials become significant. Sources of 
variation include the amount of epoxy and 
the concentricity of the optical fiber to the 
ferrule. Conditions such as these increase 
the variability of the thermal time constant 
resulting in greater function time jitter. 

This has system level implications. The 
specification for the laser diode firing unit 
pulse duration needs to balance the output 
power of presently available laser diodes 
versus the inherent increase of non- 
repeatability of function time as the pulse 
duration is lengthened to allow for lower 
net laser diode power. Another way to 
state this is function time jitter will be 
minimized when the laser diode with the 
highest output power is used. Also, 
thermal isolation of the propellant from 
surrounding materials will result in more 
repeatable ignition and, therefore, lower 
all-fire power level which in turn means 
laser diode output power requirements are 
reduced. 

PRESSURE PERFORMANCE 

For many NSI users, the interest is in 
output pressure characterization since the 
NSI is used as a cartridge to actuate 
pressure driven devices (eg bolt cutters, 
pin pullers, etc). The LNSI has 
demonstrated a nominal output pressure of 
654 psi over the last several lots of 
LNSI's; within each lot, the Coefficient of 
Variation has ranged from 3 to 5%. 



- 94 - 



Also of concern is the response time and 
time to peak pressure. At 800 milliwatts 
net power applied at the opto-propellant 
interface, function times (i.e. time from 
application of laser diode power to first 
pressure) has been 1.5 milliseconds with a 
Coefficient of Variation of 12%; time to 
peak pressure has been 0.13 milliseconds 
with a Coefficient of Variation of 20%. 

This data is difficult to compare with the 
NSI specification since requirements are 
tied to the applied current level. What can 
be said first of all is the energy source, 
whether it be a hot bridge wire or a laser 
diode, does not effect the pressure 
performance of the NSI propellant. 
Secondly, the pressure rise rate to 525 psi 
can be inferred to be equal to or better 
than the rise rate performance reported 
above for attaining peak pressure. 
Further, all of the NSI pressure-time 
requirements at 3.5 amps can be met: 

(1) Time to first pressure of greater 
than 1.0 milliseconds 

(2) Time to 525 psig shall not exceed 
6.0 milliseconds 

(3) Peak pressure shall be 525 to 775 
psig 

(4) Range of time to first pressure shall 
not exceed 3.5 milliseconds 

(5) Range of pressure rise from first 
pressure to 525 psig shall not 
exceed 0.5 milliseconds 



Random Vibration will consist of 1 
min/axis exposures to the level indicated 
on Figure 3 . 

Thermal cycle exposure will be from 
-16°C to +56°C (3°F to 133°F) for a 
total of 6 cycles with a 2 hours minimum 
dwell at temperature. 

The final two test groups will be used in 
reliability tests to establish all-fire power 
at hot and cold temperatures. 

RECOMMENDATIONS FOR FUTURE 
DEVELOPMENT 

Further development of the LNSI is 
needed to expand the performance 
envelope. No problems are anticipated 
from vibration due to the mechanical 
similarities between the NSI and the LNSI. 
The thermal environment does however 
raise some questions regarding low 
temperature performance of optical fibers. 

A parallel task is to develop a specification 
for an LNSI. Specific areas needing 
attention are: (1) all-fire power and pulse 
width, (2) no-fire power and pulse width 
(which also requires credible stray light 
sources to be identified and quantified) 
and, (3) pressure versus time performance. 



PLANNED TESTING 

Development testing is on-going at Ensign- 
Bickford. The next test series includes 
exposure to random vibration and thermal 
cycling along with high and low 
temperature all-fire power reliability tests. 
The requirements were customer driven 
based upon satellite requirements. The 
test matrix is shown in Table 1. 



- 95 - 



TTDd© Ensign-Bickford 



LASER WELD 



3 7 5-24 UNJ F - 3A 






OPT I CAL 
FIBER 




OPT I CAL 
HEADER 



LASER WELD 

PETAL L I NG 
CLOSURE D I SC 



NS I 

PROPEL L ANT 



SOU I B 
HOUS I NG 



TABLE 1-- FLIGHT READINESS TEST MATRIX 



CO 
00 



Itest 






GROUP 








A 

9 Units 


B 

9 Units 


C 
9 Units 


D 
9 Units 


E 

20 Units 


F 
20 Units 


INSPECTION 


• 


• 


• 


• 


• 


• 


THERMAL 
CYCLE 






• 


• 






RANDOM 
VIBRATION 




• 




• 






FUNCTION 
High Temp 
Ambient 
Low Temp 


3 
3 
3 


3 
3 
3 


3 
3 
3 


3 
3 
3 


20 


20 



Figure 2 

LASER INITIATED NSI PERFORMANCE 

200ji Optical Fiber Pigtail 
Room Temperature Data 




Function Time (msec) 



Figure 3 — Random Vibration Spectrum 






Q 
u 

& 
L. 

O 
On 



U. 1 


























































































H . 






















































H 


I 
























































\ 
















.... 








































\ 






















































\ 
























































\ 
















■- 








































\ 
















n ni 






































\ 


1 
















U.U1 




































































































































































































































































































































































































... 
























































0.001 - 

























































10 



100 



1000 



10000 



Frequency (Hz) 



si- n 



U^ 



Four Channel Laser Firing Unit Using Laser Diodes 

David Rosner, Sr. Electrical Development Engineer 

Edwin Spomer, Sr. Electrical Development Engineer 

Pacific Scientific/ Energy Dynamics Division 



ABSTRACT 

This paper describes the accomplishments and 
status of PS/EDD's internal research and devel- 
opment) effort to prototype and demonstrate a 
practical four channel laser firing unit (LFU) 
that uses laser diodes to initiate pyrotechnic 
events. The LFU individually initiates four ord- 
nance devices using the energy from four diode 
lasers carried the over fiber optics. The LFU 
demonstrates end-to-end optical built in test 
(BIT) capabilities. Both Single Fiber Reflective 
BIT and Dual Fiber Reflective BIT approaches 
are discussed and reflection loss data is present- 
ed. 

This paper includes detailed discussions of the 
advantages and disadvantages of both BIT ap- 
proaches, all-fire and no-fire levels, and BIT 
detection levels. The following topics are also 
addressed: electronic control and BIT circuits, 
fiber optic sizing and distribution, and an 
electromechanical shutter type safe/arm device. 
This paper shows the viability of laser diode 
initiation systems and single fiber BIT for typi- 
cal military applications. 



1. INTRODUCTION. 

1.1 Purpose . This paper presents the accom- 
plishments and status of Pacific Scientif- 
ic/Energy Dynamics Division (PS/EDD) internal 
research and development effort to prototype 
and demonstrate a practical Four Channel Laser 
Firing Unit (LFU) incorporating laser diodes. In 
this program, PS/EDD is developing and de- 
monstrating laser diode initiated safe/ami tech- 
nology for commercial, space, and defense ap- 
plications. 

1.2 Design Goals . PS/EDD designed the LFU 
as a Safe and Arm Device (SAD) shown in 



Figure 1 for a typical missile application requir- 
ing flight functions like stage separation, motor 
ignition, and shroud removal. We designed it to 
operate in typical missile environments. 
EMI/EMP protection features, and systems for 
built-in-test (BIT) of the optical and electronic 
subsystems were incorporated. We chose a sim- 
ple electronic interface using redundant elec- 
tronic controllers that can be tailored to support 
a more sophisticated interface. 

As much as practical, we designed the LFU to 
address the typical Military safety specifications 
and guidelines for in-line SAD. The LFU uses 
one electromechanical energy barrier in the opti- 
cal path, several static switches in the arm and 
firing circuits, and no mechanical barrier in the 
ordnance train. 

1.3 Typical Safety Requirements . Ordnance 
subsystems must often meet certain documented 
safety criteria. The following documents are 
some of the specifications that can be applied to 
ordnance subsystems in military and aerospace 
systems: 

• MIL-STD-1316D titled "Military Standard, 
Fuze Design, Safety Criteria For" 

• MIL-STD-1512 titled "Military Standard, 
Electroexplosive Subsystem, Electrically 
Initiated, Design Requirements and Test 
Methods" 

• MIL-STD-1576 titled "Military Standard, 
Electroexplosive Subsystem Safety Require- 
ments and Test Methods For Space Sys- 
tems" 

• MIL-STD-1901 titled "Military Standard, 
Munition Rocket and Missile Motor Ignition 
System Design, Safety Criteria For" 

• WSERB Guidelines Titled "WSERB Tech- 
nical Manual For Electronic Safety and 
Arming Devices with Non-Interrupted Ex- 
plosive Trains" 



- 101 - 



3.50 



J2 Fiber Optic 
Connector 



J1 Control & Power 
Connector 




Figure 1. Four Channel Laser Firing Unit 



Both MIL-STD-1316D and MIL-STD-1901 
address interruption type and in-line ordnance 
subsystems. The WSERB Guidelines specifical- 
ly address in-line ordnance systems and are of- 
ten specified in addition to MIL-STD-1316D 
and MIL-STD-1901. However, these specifica- 
tions do not directly address laser initiation sys- 
tems. The LFU is designed to address these 
requirements and guidelines as much as practi- 
cal. 



2. LFU OVERVIEW. 

2.1 Introduction . The LFU uses laser diodes 
and solid state electronics to initiate ordnance 
devices via fiber optics. It contains a Laser Di- 
ode Safe/Arm Module (LDSAM) to provide the 
safety -reliability of a movable barrier or shutter. 
It also contains a built-in test system that per- 
forms an end-to-end test of the fiber optic.paths. 

The size of the LFU shown in Figure 1 is 

6.80 in. x 4.25 in. x 3.50 in. and it wfeighs 3.8 lb 

excluding connectors and cables. The LFU is 



equipped with MIL-C-38999 Class IV connec- 
tors for both electrical and fiber optic interfaces. 
The input power requirements are 28 Vdc at 
0.4 Adc average and 3.6 Apeak for 10 ms when 
firing a laser. 

The input commands enter the LFU through Jl. 
Each uses an opto-isolated pair of connections 
that can be driven by 5 V TTL logic. The input 
commands are: 

• Master Reset Command resets the LFU log- 
ic and starts operation in the selected 
Test/Launch mode. This is essentially a 
powerup reset without cycling power. 

• Test/Launch Mode Command instructs the 
LFU either to perform an automatic BIT 
sequence (Test) or to execute the normal 
operational sequence (Launch) after power- 
up or master reset. 

• Pre-arm Command energizes the Pre-arm 
Switch Circuit to make electrical power ava- 
ilable to the electronic switches for the high 
power laser diodes and LDSAM arming 
solenoid. 



- 102 - 



Arm Command energizes the Arm Switch 
Circuit to power the LDSAM arming sole- 
noid using power available via the Pre-arm 
Switch Circuit. 

Select 1, Select 2. and Select 3 Commands 
act as a three bit code to select between the 
four LFU outputs before each fire 
command. 

Fire Command commands the LFU to sup- 
ply the high power laser pulse from the se- 
lected output provided that the LFU is 
armed. 

The two output status signals, BIT Pass and BIT 
Fail, exit the LFU through Jl. Each uses an 
opto-isolated pair of connections providing an 
electronic switch closure. 



The four laser outputs exit the LFU through J2. 
Each uses 100 nm core 0.37 numerical aperture 
(NA) fiber and provides a 904 nm wavelength 
10 ms pulse of approximately 1.0 watt. 

2.2 LDSAM Characteristics . The LDSAM is 
a 24 output electromechanical shutter assembly 
with only four outputs fully assembled. It pro- 
vides an optomechanical safety feature for ord- 
nance initiation power. As show in Figure 2, all 
critical optical elements in the LDSAM are rig- 
idly mounted to eliminate misalignment in harsh 
environments. 



Shutter Aperture 



High Power Laser 



Collimating Lens 




Alignment Collar 



Fiber Optic 



Focusing Lens 



Figure 2. LDSAM Cutaway View 



- 103 - 



For each output, a rigidly mounted laser diode 
and collimating lens illuminates a rigidly mount- 
ed focusing lens and 100 ^m core fiber optic ca- 
ble. An aluminum shutter, located between the 
collimating and focusing optics, acts as an ener- 
gy barrier. A solenoid rotates the shutter be- 
tween Safe and Arm positions with a spring 
loaded return to Safe. 

An electro-optical sensor monitors the Safe po- 
sition of shutter. Visual indication is provided 
by a shutter driven flag and window. The flag is 
labeled "S" for safe and H A H for Armed. A re- 
movable safing pin locks the shutter into the 
Safe position when installed. 

2.3 Built-in-Test (BIT) System . The LFU also 
contains a BIT system that performs the follow- 
ing tests on the LFU subsystems; 

Continuity of fiber optic paths between the 

LFU and ordnance devices 

Firing of high power laser diodes 

Operation of Pre-arm circuits 

Operation of LDS AM. 

To develop an optimal design, PS/EDD investi- 
gated two different approaches to performing 
the optical continuity BIT. The LFU is 
equipped with two channels of each type. They 
are: 

Single Fiber Reflective BIT . A low power 
laser signal is sent to the ordnance device 
through the same fiber optic used to initiate 
that device. This signal reflects off the di- 
chroic mirror deposited on the window in 
the ordnance device and returns to the BIT 
system through the same fiber optic. A pho- 
todiode and electronic circuit measure its 
intensity. 
• Dual Fiber Reflective BIT . A low power 
laser signal is sent to the ordnance device 
using the same high power diode laser and 
fiber optic used to initiate that device. The 
output power of the high power laser diode 
is limited to a level safely below the no-fire 
level by an aperture in the shutter. A small 
fraction of this signal reflects off the ord- 
nance device window and returns to the BIT 
system through a second fiber optic. The 
BIT system uses a photodiode and electronic 



circuit to measure the intensity of this refle- 
ction. 

2.4 Electronic Subsystem . The Electronic 
Subsystem controls and sequences the BIT fea- 
tures and the laser initiation system. It also in- 
terfaces to other missile systems. The Electron- 
ic Subsystem consists of the following elements 
shown in the LFU Functional Block Diagram 
(Figure 3): 

• Input and Output Circuits 

• Power Converters 

• Redundant Controllers 

• Pre-arm Switch Circuit 

• Arm Switch Circuit 

• High Power Laser Drive Circuit 
BIT Laser Drive Circuit 

BIT Sense Circuits 

2.4.1 Input and Output Circuits . We used 
optocouplers and transient suppressors for all 
electronic input and output (I/O) signals. Each 
input command enters the LFU through Jl and 
uses an opto-isolated pair of connections that 
can be driven by 5 V TTL logic. 

Each output signal exits the LFU through Jl and 
uses an opto-isolated pair of connections provid- 
ing electronic switch closure. All electrical in- 
puts and outputs, including power, are equipped 
with semiconductor transient suppressors to 
protect against electrostatic discharge (ESD) and 
electromagnetic pulse (EMP). 

2.4.2 Power Converters . The LFU is equipped 
with two DC/DC power converters that provide 
regulated sources of 5 Vdc for logic circuits, and 
of 15 Vdc for the analog circuits and the BIT 
lasers. Unregulated 28 Vdc provides power to 
both DC/DC converters and to the Pre-arm Cir- 
cuit. Note, the DC/DC Converters operate when 
28 Vdc is present no matter whether the Pre-arm 
Switch Circuit is open or closed. 

The power returns for the 28 Vdc, 5 Vdc, and 
15 Vdc sources are separately routed. They are 
connected to the chassis at a single point 
through 47 kfl resistors bypassed by 0.01 nF 
capacitors. This minimizes cross talk between 
the digital, analog, and laser fire circuits. 



- 104 - 



Channel Select Inputs 

Pre-Arm Command 

Arm Command 

Master Reset 

Launch/BIT Select Input 




p/o Jl 



BIT Pass Status 

BIT Fail Status 



Fiber Optic 
Outputs to 
Ordnance 
Devices 



Channel 1 Output 
(Single Fiber BIT) 



Channel 2 Output 
(Single Fiber BIT) 



Channel 3 Output 
(Dual Fiber BIT) 



Channel 4 Output 
(Dual Fiber BIT) 



Figure 3. LFU Functional Block Diagram 



2.4.3 Redundant Controllers . PS/EDD chose 
to use hard logic implemented with application 
specific integrated circuits (ASIC) instead of 
using microprocessors or stored program devic- 
es. This choice eliminates the costs involved in 
developing, debugging, and eventually qualify- 
ing software. 

Each Redundant Controller consists of a single 
ACTEL brand factory programmable logic array 
(FPGA). Both controllers are identical and op- 
erate from a common clock. A common power- 
up reset circuit resets the logic circuits in each 
FPGA and initiates the automatic BIT testing of 
the LFU. 



The BIT logic circuit part of each FPGA is basi- 
cally a string of latches fed by logic gates. This 
forms a state machine that sequences through a 
set of predefined steps. Each step sets the 
FPGA's outputs to predefined levels and per- 
forms a boolean logic test of all inputs. If the 
test passes then the logic proceeds to the next 
state, if it fails the logic indicates a BIT failure 
and stops. 

The fire control logic circuit part of each FPGA 
consists of logic circuits to decode the channel 
selection and fire commands originating form 
outside the LFU. It also consists of timing cir- 
cuits to control the duration of the High Power 
Laser outputs. 



105 - 



Each FPGA monitors the following commands 
originating from outside the LFU: Master Reset 
Command; Test/Launch Mode Command; Pre- 
arm Command; Arm Command; Channel Select 
Inputs (Select 1, Select 2, and Select 3 Com- 
mands); and Fire Command. Each FPGA also 
monitors the following internal signals: Pre-arm 
Switch Circuit monitor; Arm Switch Circuit 
monitor; Safe position status of LDSAM; and 
output of the Bit Sense Circuit. 

Each FPGA generates the Bit Pass and Bit Fail 
status commands for use by missile systems 
outside the LFU. Each also generates com- 
mands to arm the Arm Switch Circuit and to fire 
High Power Laser Diode Drive Circuits. The 
FPGA's command outputs are ANDed together 
by the subsystems they control. In other words, 
both FPGAs must issue identical output com- 
mands before a subsystem uses that output 
command. The BIT Fail status outputs are 
ORed together so that either controller can issue 
a BIT failure signal. 

2.4.4 Pre-arm and Arm Switch Circuits . The 

LFU uses unregulated 28 Vdc to energize the 
LDSAM solenoid and to power the High Power 
Laser Diodes. The unregulated power is first 
routed through the Pre-arm Switch Circuit to 
provide the first static switch function for arm- 
ing and ordnance initiation power. 

The Pre-arm Switch Circuit uses MOSFET 
switches to switch both the +28 Vdc and 28 Vdc 
return lines, and is commanded by the Redun- 
dant Controllers through two series connected 
optocouplers. This arrangement requires that 
both Redundant Controllers issue the Pre-arm 
command to turn on the Pre-arm Switch Circuit. 
The Pre-arm Switch Circuit also provides a sin- 
gle monitor signal to both Redundant Control- 
lers through an optocoupler. 

The switched +28 Vdc and 28 Vdc return out- 
puts of the Pre-arm Switch Circuit is then routed 
to the High Power Laser Drive Circuits and to 
the Arm Switch Circuit. 



MOSFET switches to control both the +28 Vdc 
and 28 Vdc return lines, and to energize the 
LDSAM solenoid. It is also commanded by the 
Redundant Controllers through two series con- 
nected optocouplers and provides a single moni- 
tor signal to both Redundant Controllers through 
an optocoupler. 

2.4.5 High Power Laser Drive Circuit This 
drive circuit consists of four individual 
MOSFET switches to control each of the four 
High Power Laser Diodes. These MOSFET 
switches receive switched +28 Vdc power from 
the Pre-arm Switch Circuit through a common 
current regulator circuit. The current regulator 
compensates for variations in the unregulated 
28 Vdc power source and provides a constant 
current to the High Power Laser Diodes. It is 
designed to operate the lasers within their rated 
power limits. 

The MOSFET switches provide the second stat- 
ic switch function for ordnance initiation power 
and are controlled by the Redundant Controllers 
through two series connected optocouplers. As 
with the Pre-arm and Arm Switch Circuits, this 
arrangement requires that both Redundant Con- 
trollers issue the fire command to turn on a laser 
diode. 

2.4.6 BIT Laser Drive Circuits . This drive 
circuit consists of two individual MOSFET 
switches to control each of the two BIT Laser 
Diodes used in the single fiber BIT system. 
These MOSFET switches receive regulated 
15 Vdc from the Power Converters through a 
common current regulator circuit. The current 
regulator provides a constant current to the BIT 
Laser Diodes and operates the lasers within their 
rated power limits. 

The MOSFET switches are controlled by the 
Redundant Controllers through two series con- 
nected optocouplers. As with the High Power 
Laser Drive Circuits, this arrangement requires 
that both Redundant Controllers issue the fire 
command to turn on a laser diode. 



The Arm Switch Circuit provides the second 
static switch function for arming power. It uses 



2.4.7 BIT Sense Circuit This circuit supports 
the optical continuity BIT function by sensing 



- 106 - 



the reflected light and by providing a simple 
digital signal to Redundant Controllers. The 
BIT Sense Circuit supports both Single and Dual 
Fiber BIT systems, and consists of: 

Photo diodes to sense the reflected BIT sig- 
nals 
• Analog circuits for amplification and level 
detection 

Optocouplers for output to the Redundant 
Controllers 

The sensitivity of the BIT Sense Circuit is limit- 
ed by the photo diode's rated dark current while 
the response time is limited by the photo diode's 
total capacitance rating. In other words, the 
optical BIT signal must be bright enough to be 
reliably detected above the photo diode's worst 
case dark current and must be present long 
enough for the photo diode's capacitance to 
charge up. 



3. LASER DIODE INITIATION SUBSYS- 
TEM. 

3-1 Introduction . The basic mechanism for 
laser ignition is thermal in nature. The laser 
ignition system must deliver a sufficient intensi- 
ty to raise the temperature of the ordnance com- 
pound above its ignition temperature. This de- 
pends on the properties of the ordnance com- 
pound such as: ignition temperature, thermal 
diffusivity, specific heat, surface optical 
properties, and particle size. 

The Laser Diode Initiation Subsystem uses con- 
tinuous type lasers that are typically rated in 
watts. For this reason, it is best to specify the 
all-fire and no-fire levels in units of power 
(milliwatts) instead of units of energy 
(millijoules). Since the all-fire and no-fire lev- 
els depend on spot size, the type of fiber used to 
deliver the laser energy to the ordnance device 
must be specified. 

For this design we used a 100 ^m core step in- 
dex fiber optic inside the LFU and 1 10 ^im core 
step index fiber for the external cables. This 
conserves the intensity of the laser diode as it is 
delivered to the ordnance device. 



The optical path starts with a 920 nm wave- 
length High Power Laser Diode coupled to a 
collimating lens and mounted in the LDSAM, 
shown in the Initiation Subsystem Functional 
Block Diagram Figure 4. The collimated laser 
light passes through the LDSAM shutter and is 
refocused into a 100 nm fiber using another lens 
in the LDSAM. For channels one and two, the 
coupler is the next item in the optical path. Fi- 
nally the J2 connector on the LFU is the last 
item in the optical path. Table 1 lists the typical 
output delivered to an ordnance device through 
the external 110 nm cables. 

3.2 Requirements . There are no established 
industry-wide all-fire and no-fire standards for 
laser ordnance. However, PS/EDD has designed 
several diode initiated devices. As an example, 
we designed and manufactured a miniature pis- 
ton actuator that had a 320 raw all-fire and a 
130 mw no-fire using 1 10 ^m diameter fiber. 
We used these levels as a guide in developing 
the LFU. 

This actuator used titanium potassium 
perchlorate for the ignition/output charge and 
incorporates a fiber optic pigtail. A Neyer sta- 
tistical analysis was used to determine the all- 
fire and no-fire levels. Table 2 shows the test 
data from the Neyer test performed on the piston 
actuator. The power levels listed in the table 
were measured prior to each test shot using the 
fiber that connects to the pigtail of the device. 

3.3 Design Trades . The main goal in designing 
the LFU is to maximize the efficiency of deliv- 
ering power to the ordnance devices. This usu- 
ally requires selecting a fiber optic cable with 
the smallest diameter practical and usually be- 
comes a trade between launch efficiency and 
spot size. In other words, one must select a 
combination of laser diodes and fibers that sup- 
ply the largest power per unit area. Minimizing 
the quantity of connector interfaces is another 
design goal. 

In designing ordnance devices, the main goal is 
to minimize the spot size at the ordnance com- 
pound while meeting other requirements like 
cost, proof pressure, and sealing. Fiber optic 



- 107 - 



Table 1. Measured LFU Output 



Output 
Channel 


Typical 
Output 
Power 


Margin 
(based on 
320 mw all- 
fire) 


1 


27.1 dBm 
(517 mw) 


162% 


2 


26.6 dBm 
(452 mw) 


141% 


3 


27.0 dBm 
(500 mw) 


156% | 


4 


26.8 dBm 
(482 mw) 


151% 



cables emit light in a diverging cone that causes 
the spot size to grow with distance. To mini- 
mize the spot size, the ordnance device must 
either put a fiber in contact with the ordnance 
compound or use optics to refocus the spot onto 
the ordnance compound. Plane parallel win- 
dows are not typically used in devices initiated 
by laser diodes. However, a gradient index 
(GRIN) lens or an integral fiber can efficiently 
couple a fiber's output to the ordnance com- 
pound. 



3.4 Outnut Tests . We measured the LFU out- 
puts at the ordnance end of external 1 10 nm 
cables. The results are listed in Table 1. Note, 
the LFU can delivers sufficient laser intensity to 
initiate ordnance devices with 141% to 162% 
margins above an 320 mw all-fire requirement. 
This margin means that the LFU operates from 
4.3 to 6.4 times the 30.7 mw sigma over the 
0.999 reliability all-fire level shown in Table 2. 

Table 2. Neyer Analysis of Laser Initiated Pis- 
ton Actuator 



Stimulus in 
milliwatts 


Successes 


133.0 





172.0 





190.0 





213.0 


1 


241.0 





243.5 




279.0 




289.0 




298.0 




460.0 





Note: 

The Mu was 222.9 mw with a sigma of 
30.7 mw. The calculated 0.999 all-fire 
level is 317.8 mw and the 0.001 no-fire 
level is 128.1 mw. 



- 108 - 



r 



Laser Firing Unit 



l 



High Pwr 

Laser Drive 

Circuit 



High Pwr 

Laser Drive 

Circuit 



fO 



Safe 




30 



Channel 1 



Channel 3 



Channel 4 

(same as Channel 3) 



Shutter in 
Arm T u Arm Position 

Laser Diode S/A Module 



l 

Figure 4. Initiation Subsystem Functional Block Diagram 



J2 



Connector at 
Ordnance Device 



All-Fire 
Power 



Connector at 
Ordnance Device 



All-Fire 
Power 



J 



4. SINGLE FIBER BIT SUBSYSTEM. 

4.1 Introduction . The Single Fiber BIT Sub- 
system is meant to detect broken fiber optics or 
mismated and contaminated optical connections. 
Each Single Fiber BIT channel consists of fol- 
lowing elements shown in Figure 5: 

• A 1.0 mw BIT Laser Diode operating at 
780 nm wavelength 

A fiber coupler that has three inputs and one 

output 

A common BIT Sense Circuit to detect the 

reflected optical BIT signal. 

• An ordnance device with a dichroic coating 
on the window that reflects 780 nm wave- 
length light. 

The Optical Coupler provides paths for injecting 
the BIT laser signal into the output fiber and for 
extracting the return signal from the output fi- 
ber. The three inputs of the coupler are connect- 
ed to the High Power laser diode, BIT laser di- 
ode, and the BIT photodiode. 



During Single Fiber BIT operation, the LDSAM 
is in the safe position with the output of the 
High Power Laser Diodes blocked by the closed 
LDSAM shutter. A BIT laser is fired to provide 
a low power laser pulse to illuminate the ord- 
nance device through the single output cable. 
The dichroic coating on the ordnance device 
window reflects 90% of the BIT laser output 
back to the LFU through the same cable. The 
fiber coupler directs this reflected signal to the 
BIT Sense Circuit. 

4.2 Requirements . To keep with the spirit of 
the monitor circuit requirements of MIL-STD- 
1516, the power of the BIT signal should be 
kept 20 dB below the rated no-fire of the ord- 
nance device. Note, MIL-STD-1516 para- 
graph 5.10.7 limits the monitor current for 
electroexplosive devices to one tenth of the no- 
fire level. This corresponds to a one hundredth 
factor for power or a margin of 20 dB. 



- 109 - 



F 



BIT 
Sense 
Circui 



7> 

it —J 



BIT Laser 

[ 
Ci 



IT Laser "IL 
Drive \l 
Circuit — ' 



780 nanometer 



High Pwr 

Laser Drive 

Circuit 



920 nanometer 



L 



Laser Firing Unit 



~i 



J 



Safe 



I V Optical 

I ^^ 4 ^Coupler 



J2 



- Shutter in 
Safe Position 



Arm T 

Laser Diode S/A Module 



SMA Connector 



Connector at 
Ordnance Device 



J 



Figure 5. Single Fiber BIT Subsystem Functional Block Diagram 



The BIT Laser Diodes are rated at 1.0 mvv 
which is 21 dB below the no-fire level of 
130 mw. This 1.0 mw output is attenuated by 
the 1 1 dB loss in the coupler and the coupling 
loss to the fiber. In other words, the laser inten- 
sity at the ordnance device is at least 31 dB be- 
low the no-fire level of 130 mw or 12 dB lower 
than the MIL-STD-1516 requirement. Note, this 
does not include the losses associated with the 
connectors (typically 0.7 dB to 1.5 dB loss per 
connector) and with the dichroic coating (10 dB 
loss i.e. it passes 10% at 780 nm). 

4.3 Design Trades . Unwanted reflections are 
the main limiting factor for the Single Fiber BIT 
system. These are caused by the connectors in 
the optical path and by the coupler's internal 
reflections. These reflections are sensed by the 
BIT Sense Circuit and appear as background 
noise that the BIT reflection must over come. 
For connectors, the fresnel reflection at each 
glass-to-air interface is 4.0% of the incident 
light. The intensity of the coupler's internal 
reflections are approximately 25 dB below (or 
0.32% of) the BIT Laser intensity. The main 
design trade is to optimize the coupler design to 
minimize reflections while still providing an 
adequate optical path for initiation energy. 



4.4 BIT Tests . We performed some testing to 
detenu ine the reflection losses that we can ex- 
pect at the ordnance device during Single Fiber 
BIT. A mirror and GRIN lens simulated the 
ordnance device with dichroic coating. A 
1 10 um fiber optic with an SMA connector was 
routed from the LFU and mated with the GRIN 
lens/mirror combination. Using an optics bread- 
board we could vary the gap distance between 
the SMA connector and the GRIN lens. Using a 
HeNe laser in place of the BIT laser diode, we 
measured the reflection loss for a variety of 
gaps. Figure 6 is a plot of the relative loss vers- 
es gap distance. The loss values are referenced 
to the BIT output intensity of the LFU. 

As Figure 6 shows, there is a 1.5 dB dynamic 
range for discriminating between pass and fail. 
However the intensity of the reflection is only 
about 22 dB to 24 dB below the LFU's BIT out- 
put. In the spirit of MIL-STD-1516, the BIT 
reflection could be as bright as 44 dB below the 
rated no-fire of the ordnance device. This could 
be as much as 5.2 jiw for a system using 
130 mw no-fire ordnance devices. Silicon 
photodetectors have a typical sensitivity of 
0.5 nA/jiw and would generate a 2.6 \iA signal 
that is easily detectable. 



- no - 



-22 





-22.2 


CD £ 




"D ti 




</> o 


-22.4 


</) 




O 3 

—I u. 


-22.6 


sz -> 




"co 2 


-22.8 


C CD 




.2 g 


-23 


o <u 




© a5 


-23.2 


a> <5 




a: oc 






-23.4 




-23.6 




0.00 100 2.00 3.00 4.00 

Gap between Fiber & GRIN/Relfector Combination (mm) 



5.00 



Figure 6. Reflection Loss Verses Gap Distance for Single Fiber BIT 



5. DUAL FIBER BIT SUBSYSTEM. 

5.1 Introduction . The Dual Fiber BIT Sub- 
system is meant to detect mismated and contam- 
inated optical connections. It uses two separate 
fibers from the LFU that are terminated together 
in the connector that mates with the ordnance 
device. One fiber connects to a LDSAM output 
while the other connects to a PIN diode 
photodetector in the BIT Sense Circuit as shown 
in Figure 7. 

During BIT, the LDSAM is in the safe position 
and a high power laser diode is fired to provide 
a low power laser pulse to illuminate the ord- 
nance device. Note, the output power of the 
high power laser diode is limited by an aperture 
in the shutter. A small fraction of this signal 
reflects off the ordnance device and returns to 
the BIT Sense Circuit through a second fiber op- 
tic. 

We wanted to minimize costs by making the 
ordnance interface simple fabricate. The cable 
interface at the ordnance device consists of an 



SMA type fiber optic connector with two fibers 
bonded side-by-side and polished. The fiber 
core diameter was 1 lO^im with an overall diam- 
eter of 125 nm, and a numerical aperture (NA) 
of 0.37. 

5.2 Requirements . To keep with the spirit of 
the monitor circuit requirements of MIL-STD- 
1516, the power of the BIT signal should be 
kept 20 dB below the rated no-fire of the ord- 
nance device. Note, MIL-STD-1516 para- 
graph 5.10.7 limits the monitor current for elect- 
roexplosive devices to one tenth of the no-fire 
level This corresponds to a one hundredth fac- 
tor for power or a margin of 20 dB. 

The High Power Laser Diodes have a 13 dB 
dynamic range from 0.1 W at threshold to a 
maximum of 2.0 W. This dynamic range is not 
large enough to use current limiting alone to 
reach the 20 dB safety margin. Therefore a 
shutter with either an aperture or other type of 
optical attenuator was required. We found that 
an aperture of about 0.005 in. provides 27 db to 
30 dB of attenuation. 



- in - 



r 



Laser Firing Unit 



High Pwr 

Laser Drive 

Circuit 



£ 



\ 



BIT 
Sense 
Circuit 



■> BIT Aperture 



- Shutter in 
Safe Position 



Arm T 

Laser Diode S/A Module 



J2 



3q 



SMA Connector 



Connector at 
Ordnance Device 



I 

Figure 7. Dual Fiber BIT Subsystem Functional Block Diagram 



5.3 Design Trades . The main design trade for 
a Dual Fiber BIT systems is cost verses the in- 
tensity of reflected signal. For example, we 
could have used a coupler, a separate BIT laser 
operating at 780 nm, and a dichroic coating on 
the ordnance device. This would have increased 
the reflected signal however the system cost and 
configuration are very similar to the Single Fi- 
ber BIT configuration. Alternately, we could 
have complicated the ordnance interface to im- 
prove the intensity of the reflected signal. How- 
ever this would significantly increase the cost of 
the ordnance devices. 

For our low cost approach, the critical design 
trade is to maximize the reflected BIT signal 
while providing adequate initiation power. 
Figure 8 shows the ray diagram associated with 
a pair of fibers up against a reflector. The re- 
flecting surface represents the ordnance com- 
pound and fresnel reflections from and imaging 
optics associated with a practical ordnance de- 
vice. Unlike the single fiber BIT approach, di- 
chroic coatings can not be used since this ap- 
proach uses the same laser for BIT and initia- 
tion. 

As mentioned in section 3.3, any laser diode 
initiated ordnance device would use optics be- 
tween the fiber and the ordnance compound to 



re-image the laser spot while providing a seal. 
Ideally, the net effect of such optics would be 
the same as not having any optics. For simplici- 
ty, Figure 8 does not show any re-imaging op- 
tics. 

The bottom fiber illuminates the reflector while 
the top fiber gathers the reflected light. Both 
fibers emit or accept light in diverging 43.4 de- 
gree cones that overlap at the reflecting surface. 
The intensity of the reflected BIT laser signal is 
directly effected by this overlap area and by 
reflectivity of the ordnance compound. 

Note that the area of this overlap varies with the 
gap between the reflector and the fiber ends. 
For a given set of fiber core diameters and fiber 
spacing, there is a range of usable gaps with 
specific minimum and maximum gap distances. 
One can expect the intensity of the reflected BIT 
to increase with gap distance to a peak value and 
then decrease with larger gap distances. To 
avoid the ambiguity of having an intensity value 
indicate two different gap distances, a designer 
would select a gap that is ate or beyond the 
peak. 



- 112 - 



.110 mm core dia 

0.125 mm overall dia 

NA=0.37 

Half angle - 21.72 deg 




Acceptance & 
radiation angle 
determined by NA 



Reflecting surface 



Acceptance & 
radiation angle 
determined by NA 



Usable range for 
Reflecting surface 



Figure 8. Dual Fiber BIT Sense Geometry 



As mentioned in section 3.1, any gap in the 
fiber-to-ordnance interface increases spot size 
requiring more all-fire power from the High 
Power Laser Diodes. With this type of fiber 
geometry, one must select the BIT laser, fiber 
size, and ordnance interface geometry to maxi- 
mize the reflected BIT signal while providing 
adequate initiation power for the resulting spot 
size. 

5.4 BIT Tests . We performed some testing to 
determine the reflection losses that we can ex- 
pect at the ordnance device during Dual Fiber 
BIT. A flat black surface and GRIN lens simu- 
lated the ordnance device. A pair of 1 10 \xm 
fibers were terminated in an SMA connector 
mated with the GRIN lens/black surface combi- 
nation. One fiber was routed from a HeNe laser 
to simulate the LFU's BIT output and the other 
fiber was monitored by an optical wattmeter. 

Using an optics breadboard we could vary the 
gap distance between the SMA connector and 
the GRIN lens. Figure 9 is a plot of the relative 
loss verses gap distance. The loss values are 
referenced to the output intensity of the LFU. 
The loss curve starts at a low level, peaks at - 
25 dB, and the gradually drops to -40 dB. Note 
that losses between -25 dB to -40 dB correspond 
to two different gap values. 



As discussed in section 5.3, the selected fiber 
geometry has a range of usable gaps with specif- 
ic minimum and maximum gaps. The effect of 
the minimum gap can be seen in the steep rise 
while the effect of the maximum gap is seen in 
the curve's fall. 

However, one would expect a steeper fall than 
shown. We believe that the gradual drop is 
caused by reflections from the test setup. Simi- 
lar reflections might be expected from a connec- 
tor that is partially mated to an ordnance device 
with its clean reflective connector interface. In 
any case, the such reflections effect the sensitiv- 
ity of this BIT approach. 

As Figure 9 shows, there is a 15 dB dynamic 
range for discriminating between pass and fail. 
However the intensity of the reflection ranges 
from 26 dB to 40 dB below the LFU output. In 
the spirit of MIL-STD-1516, the BIT reflection 
could be as bright as 46 dB to 60 dB below the 
rated no-fire of the ordnance device. This could 
be as much as 3.3 nw to 130 nw for a system 
using 130 mw no-fire ordnance devices. Silicon 
photodetectors have a typical sensitivity of 
0.5 nA/nw and would generate a 1.7 \xA to 
65 nA signal that is not easily detectable. 



- 113 - 





f^&s?. 










co O ou 
O 3 

.c -■ 












ection P< 
fenced 1 

D C 

1 












5= OJ ^ 

0) 0) 

or rr 

-60 h 


i i I i 


i i i i 


i i i i 


i i i i 


i i i i 



0.00 



1.00 



2.00 



3.00 



4.00 



5.00 



Gap between Fiber & GRIN/Black Relfector 
Combination (mm) 



Figure 9. Reflection Loss verses Gap Distance for Dual Fiber BIT 



6. SUMMARY . 

6.1 Conclusions . With this effort, PS/EDD is 
demonstrating a viable laser diode based initia- 
tion system that uses Single Fiber BIT and a 
high degree of automatic operation. The LFU 
delivers sufficient laser power with 141% to 
162% margins above an 320 mw all-fire require- 
ment. This margin is 4.3 to 6.4 times the 
30.7 mw sigma over the 0.999 reliability all-fire 
level shown in Table 2. PS/EDD is demonstrat- 
ing a workable Single Fiber BIT System that 
requires only one fiber per ordnance device for 
both initiation and BIT. 



two different wavelengths and dichroic coatings 
on the ordnance devices. 

6.2 What's Remains . In general, PS/EDD 
plans to apply this technology to simpler and 
lower cost units. We will perform some envi- 
ronmental testing on LDSAM and BIT verifica- 
tion tests of the Single Fiber BIT system. In 
future laser diode initiation systems we will 
reselect laser diodes to take advantage of the 
newer high power lasers that have are now 
available. In future Single Fiber BIT Subsys- 
tems we will update the coupler design to fur- 
ther reduce internal reflections. 



Our aggressive approach to Dual Fiber BIT is 
proving to be unsuitable for diode based initia- 
tion systems. For simplicity, we used an 
extremely simple fiber-to-ordnance device inter- 
face that requires a gap to obtain reasonable BIT 
return signals. This gap degrades the delivery of 
initiation energy. Our approach, is better suited 
for solid state laser initiation systems that use 



- 114 - 



QMir 



LIO Validation on Pegasus 
(Oral Presentation Only) 



Arthur D. Rhea 

The Ensign-Bickford Company 
Simsbury, CT 



- 115 - 



5lO-'2fc 



EBW'S AND EFI'S 
THE OTHER ELECTRIC DETONATORS 

RON VAROSH 

RISI 



INTRODUCTION 

Exploding BridgeWire Detonators (EBW) 
and Exploding Foil Initiators (EFI) 
which were originally developed for 
military applications, have found 
numerous uses in the non-military 
commercial market while still retain- 
ing their military uses. 

While not as common as the more 
familiar hot wire initiators, EBW s 
and EFI's have definite advantages in 
certain applications. These advan- 
tages, and disadvantages, are dis- 
cussed for typical designs* 

HISTORY 

EBW 1 s were invented in the early 
1940' s by Luis Alvarez as part of the 
Manhattan project ( 1 ) . Alvarez ' s in- 
sight was to use a rapidly discharg- 
ing capacitor to fire a hot wire 
detonator and thus obtain the re- 
quired simultaneity for a nuclear 
device. Further research showed that 
this "exploding wire concept" could 
also be used to initiate secondary 
explosives such as PETN and RDX. The 
concept remained classified for many 
years until a patent was issued in 
1962 to Lawrence Johnston, one of 
Alvarez ' s co-workers . These detona- 
tors were studied and used extensive- 
ly by the former Atomic Energy Com- 
mission. Although many of these 
studies have never been declassified, 
a good sampling of what was learned 
was published in the proceedings of 
the Exploding Wire Conferences (2). 

Of particular interest in these 
conference proceedings are many of 
the reports of T.J. Tucker , especial- 
ly his formulation of the "action" 
concept (3) . The "action" concept 



remains the basis for the design and 
evaluation of both EBW's and EFI's* 

EFI ' s (or slappers as they are fre- 
quently called) were invented by John 
Stroud of Lawrence Livermore National 
Laboratory in 1965 (4) . In a report 
on the acceleration of thin plates by 
exploding foils, Stroud noted that 
the pressures produced by these 
"slappers" appeared to be sufficient 
to initiate high density secondary 
explosives* 



DEFINITIONS OF EBW'S AND EFI'S 
The same basic definition can 
applied to both EBW's and EFI's. 



be 



An EBW (or EFI ) is an all 
secondary explosive detonator 
that requires a unique, high 
ampl itude , short duration 
electrical pulse for proper 
functioning. 

Both EBW's and EFI's require a unique 
high amplitude electrical pulse, and 
each contains only secondary explo- 
sives. The differences are that the 
explosive in an EBW is directly 
against the bridgewire and is usually 
at 50% of crystal density. The 
explosive is "believed" to be shock 
initiated by the exploding wire. In 
an EFI, the exploding foil acceler- 
ates a disc across a gap and the high 
density explosive ( 90% crystal densi- 
ty) is initiated by the kinetic 
energy of the flying disc . For the 
EFI , the explosive is not in direct 
contact with the exploding foil. 

WHY BOTHER? 

These electric detonators appear to 

be so much more complicated than 



- 117 - 



PAGE 



likwiHTC:- 



simple hot wire devices, that the 
question must be addressed as to why 
bother with this added complication? 
Three major reasons why people bother 
with EBW's and EFI ' s are: 

Safety 

Repeatability 

Reliability. 

Safety comes primarily from the fact 
that no primary explosives are used 
in either device. Both are electro- 
statically safe as demonstrated by 
the "standard man test (5). Nominal 
values of RF are also not a problem 
(6). Stray currents do not affect the 
devices since approximately 3 amps DC 
are required to melt open a common 
type bridgewire and about 5 amps to 
open a typical foil. 

The second major reason for using 
EBW's and EFI ' s is their excellent 
shot to shot repeatability. Even 
with different firing systems, shot 
to shot repeatabilities under 5 
microseconds are easily obtained. 

Finally for applications where simul- 
taneity is a requirement, both EBW's 
and EFI's are easily fabricated with 
standard deviations under 25 nanosec- 
onds. 

APPLICATIONS 

Following are some of the applica- 
tions where EBW' s and EFI ' s have 
found substantial acceptance: 



nize cameras, flash X-Rays, etc. with 
detonations makes good use of the 
inherent repeatability of EBW s and 
EFI's. Much of the explosive welding 
is performed on-site to electric 
power plant boilers - a location 
notorious for stray voltages. Not 
only is the safety important here, 
but in addition many of the welds 
require the simultaneous detonation 
of two charges. Safety is the prim- 
ary requirement for the majority of 
the other applications . Mining 
applications are limited since most 
mining requires "ripple" firing - 
something extremely difficult to 
accomplish with either EBW 1 s or 
EFI's. 

EBW CONSTRUCTION 

Figure 1 shows a typical EBW detona- 
tor and compares it with a typical 




.VT77Z 



1-HEAD 
2-BRIDGEWIRE 
3-INITIAL PRESSING 
4-OUTPUT PELLET 



HOT-WIRE 

Plastic 
Hi-resist 
Lead Azide 
PETN/RDX 



E9W 

Plastic 
Lo— resist 
PETN 
PETN/RDX 



Figure 1. Typical Bridgewire Detonators 



Military Ordnance 

Military R&D 

Explosive Welding 

Explosive Hardening 

Seismic 

Oil Fields 

Forest Service 

Mining 

Power Plants 

Military Ordnance is an obvious 
application, although most conven- 
tional weapons still use hot wire 
initiators. The reverse is true with 
Ordnance R&D. The need to synchro- 



hot wire detonator. Heads and output 
pellets are the same for both detona- 
tors. The major differences are in 
the bridgewires and initial press- 
ings. EBW's generally use gold or 
platinum wires, primarily for their 
inertness, while hot wire devices use 
high resistance materials such as 
Nichrome. The explosive against the 
bridgewire in an EBW is generally 
PETN although RDX and "thermites" 
have been used . Hot wire devices 
generally have Lead Azide or Lead 
Styphnate against the bridge wire. 



- 118 - 



In addition, the explosive in an EBW 
is generally at about 50 percent of 
crystal density. In the EBW, the 
"explosion" of the wire starts a 
detonation without any intervening 
deflagration as is the case with a 
hot wire detonator. 

SAFETY DATA 

Since EBW's have been around for 
almost 50 years a great deal of test 
data has been accumulated by various 
organizations . Everyone has an opin- 
ion on which safety tests are most 
important, but the US Forest Service 
is probably the most imaginative. In 
their testing, which was conducted 
for them by China Lake (7), detona- 
tors were subjected to the following 
"potential" hazards: 

110 vac, 60 cycle 
220 vac, 60 cycle 
12 vdc battery 
truck ignition coil 
camera flash unit 
chain saw magneto 
campf ire. 

In all the above testing, all detona- 
tors dudded and none detonated . 
These are obviously hazards which 
they believe could occur in their 
work areas. 

The National Laboratories at Los 
Alamos , Sandia and Livermore have 
performed the most design studies on 
both EBW's and EFI ' s and obviously 
have the most test data on these 
devices. 



The length of the narrow section is 
approximately equal to the width i.e. 
about 0.008 inch long. 



Tamper 



Bridge Foil 



Dielectric 



Barrel 




H E Pellet 



Figure 2. EFI Major Components 



The dielectric flyer is usually 
polyimide, .001 inch thick, but other 
materials have been used. 

The barrel is usually a dielectric, 
but a conductor could also be used. 
If the diameter of the barrel hole 
equals the length of the narrow 
bridge foil length (0.008 inch), the 
design is called a "finite barrel 
design" . If the diameter of the 
barrel hole is 2+ times the narrow 
bridge foil length, the design is 
called an "infinite barrel design". 



EFI CONSTRUCTION 

Figure 2 illustrates the major com- 
ponents of an EFI . Tampers can be 
any rigid dielectric material : plas- 
tics , metals , sapphire , etc . have all 
been used successfully. Next comes 
the bridge foil. These may be any 
conductor . Copper and aluminum are 
the most common. Thicknesses are 
usually about 0.0002 inch thick. The 
thinnest width nowadays is about 
. 008 inch wide , although previously 
foils as wide as 0.025 were used. 



The above four components are lami- 
nated into one sub assembly, and 
clamped against a high density explo- 
sive pellet - usually HNS. 

In operation, for the finite barrel 
design, a high current explodes the 
narrow section of the bridge foil, 
which shears out a disc of dielectric 
which accelerates down the barrel and 
by means of kinetic energy initiates 
the high explosive pellet. An infi- 
nite barrel design works exactly the 



- 119 - 



same way except the rapid expansion 
of the dielectric "bubble" is the 
source of the kinetic energy. 

Both, EBW bridgewires and EFI foils 
"explode" because the electric cur- 
rent is heating the conductor which 
is trying to expand , but the conduc- 
tor is being heated faster than it 
can physically expand, 

FIRING CIRCUITS 

A typical firing circuit for either 
EBW s or EFI ' s is shown in Figure 
3. The circuits are similar except 



SWITCH 



r 



W BLEED < 

INPUT RESISTOR < 




L 



HIGH VOLTAGE 
CAPACITOR 



y EBW/BF1 



TRIGGER s. 
INPUT ^ 



TRIGGER 
BUFFER 



I 



such as: 

batteries 
line voltage 

piezoelectric generators 
fluidic generators 
etc. 

Most circuits also have "safety" type 
features such as bleeder resistors to 
discharge the capacitor in case of an 
aborted test , features to prevent 
repetitive firing, etc. 

TYPICAL CURRENT TRACES 

Figure 4 shows typical current traces 

through a bridgewire and a foil . 

Exploding BridgeWire 
Current Trace 




Time, microseconds 



Figure 3. Typical EBW/EFI Firing Circuit 

EFI ' s tend to use lower ca- 
pacitance values (0.1 microfarad) 
because of their requirement for low 
low inductance while EBW's usually 
use about 1.0 microfarad. The low 
inductance is necessary to "explode" 
the foil rapidly enough to accelerate 
the flyer to a high enough velocity 
to initiate the explosive. 



Exploding Foil Initiator 
Current Trace 



- Pofnt of Bu vt 



1 2 

Time, microseconds 



Figure 4. Typical Current Traces 



A wide variety of switches have been 
used for both EFI's and EBW's. These 
have included: 

overvoltage switches 
vacuum triggered switches 
gas filled triggered switches 
solid state switches 
crush switches 
etc. 

Power supplies have included systems 



Both work exactly the same way. 
Current flows through the device, 
when the bridge starts to heat the 
resistance increases and the current 
falls. At the inflection point, 
burst occurs and an arc is created. 
The arc being of lower resistance, 
allows the current to recover. 

Burst occurs when a constant action 
(integral of current squared, from 
zero to burst time) is accumulated. 



- 120 - 



Also, the current at threshold is 
constant regardless of circuit param- 
eters whereas the threshold voltage 
varies with circuit parameters. 

Burst, preferably should occur at 
about 1 microsecond for an EBW, and 
. 1 microsecond for an EFI . Longer 
times allow the wire or foil to melt 
open before exploding. 

EXPLOSIVES 

Most EBW • s use PETN which has a 
reasonable threshold firing current 
(200 amps) . RDX has a significantly 
higher threshold firing current (450 
amps) but is frequently used where 
higher operating temperatures are 
required. These threshold firing 
currents work out to be 500 and 800 
volts respectively on a 1 microfarad 
capacitor. EBW's (and EF^s) can 
also shock initiate (?) "Thermites" 
to produce an initiator with a defla- 
grating output (8). Other explosives 
tested to date with EBW's have too 
high a threshold voltage to make a 
reasonable system (above 5000 volts 
on a 1 microfarad capacitor). 



CD 

s 

o 

-C 
(0 
<D 

JZ 
H 



LU 



u 



P8X-9407 



MS 



+ 

Q 

PON 
J I I L 



J I I L 



0.1 0.2 OJ 0.4 Oi Q.8 0.7 U 0.9 

Drop Hammer Height 



Figure 5. EFI Threshold vs Drop Hammer 

MIL-STD-1316 which covers the safety 
criteria for fuzes and Safety and 
Arming Devices applies to all muni- 
tions except: 



EFI's can initiate just about any 
explosive although PETN and HNS have 
acceptably low thresholds (9). PETN 
is frequently chosen as a "weak link" 
in an explosive train because of its 
ability to sublimate away at moderate 
temperatures and dud the weapon 
system* Most DOD systems use HNS for 
two major reasons - its relatively 
low threshold and its acceptability 
by MIL-STD-1316. 

Figure 5 shows one of the main rea- 
sons why HNS is so popu 1 ar as an 
initiating explosive for EFI's. 
Plotting EFI threshold voltage versus 
a measure of explosive sensitivity 
such as Drop Hammer Height shows most 
explosives following a straight line 
(PETN-RDX-PBX9407) . HNS does not 
follow the same general trend. 
Instead it is quite insensitive as 
indicated by its large drop height, 
but still has a low threshold to EFI 
initiation. 



Nuclear Weapons 

Hand Grenades 

Flares 

Manually emplaced ordnance 

Pyrotechnic countermeasures . 

For an in-line device, the only 
permissible explosives are: , 

Comp A3 

Comp A4 

Comp A 5 

Comp CH6 

PBX 9407 

PBXN-5 

PBXM-6 

DIPAM 

HNS Type 1 

HNS Type 2 

HNS-IV. 

Note that PETN is not an acceptable 
explosive. Of the listed explosives, 
HNS appears to be the best choice for 
an in-line device. 



- 121 - 



ELECTRIC DETONATOR COMPARISON 
Table 1 compares some of the electri- 
cal characteristics of three differ- 
ent types of electric detonators. 



Hot Wire EBW 



Current 

Threshold 1 amp 
Operating 5 amps 

Voltage 

Threshold 20 volts 

Energy 

Threshold 0.2 joule 

Power 

Threshold 1 watt 

Function Time 

Typical 1 millisec. 



200 amps 
500 amps 

500 volts 

0.2 joule 

100,000 watts 
1 microsec. 



EFl 

2000 amps 
3000 amps 

1500 volts 

0.2 joule 

3,000,000 watts 
0.1 microsec. 



Table 1. Electric Detonator Comparison 

The values listed are nominal and ob- 
viously detonators have been built 
with lower energy and power require- 
ments, but the comparison is still 
useful* 

For the EBW, values assume a 1 micro- 
farad capacitor while a .15 microfar- 
ad is assumed for the EFI. The 1 amp 
1 watt hot wire device is what is 
generally required for DOD devices 
although detonators with an all fire 
current of 50 milliamps have been 
fabricated (10). 

Of particular interest, is the fact 
that all the energy values are ap- 
proximately equal. This implies that 
the same physical size fire set can 
be used for all three types of deto- 
nators. The major difference between 
the three detonators is the power. 
The higher power levels of the EBW's 
and EFI ' s are related to the very 
short energy spike associated with 
these devices. A typical bridgewire 



burst time is 1 microsecond for an 
EBW and 100 nanoseconds for an EFI. 

The equality of the energy has been 
utilized in a recently developed 
"Power Multiplier" which takes the 
energy output of a normal capacitive 
hot wire firing unit and steps up the 
power to fire an EBW (11). 

SUMMARY 

Both EBW's and EFI's tend to have 
definite advantages where safety, 
reliability and repeatability are 
required but EFI ' s have a clear 
advantage in being able to initiate 
HNS and PBX-9407 and thus are Mil- 
Std-1316 acceptable. 

As disadvantages , both tend to be 
more expensive than hot-wire blasting 
caps, but this is primarily because 
of the limited manufacturing base. 
Blasting caps are manufactured in the 
10 ■ s of millions annually while the 
total annual fabrication of EBW's is 
about 100,000. Only about 10 
20,000 EFI's are currently manufac- 
tured annually in the US. 

Two other major disadvantages of 
EBW's and EFI's, are the small number 
of detonators which can be fired per 
shot and the difficulty in delaying 
individual detonators . The ability 
to delay individual detonators is 
very important in the mining industry 
where "ripple" firing is necessary 
for efficient earth movement. 

The final disadvantage is the re- 
quirement for low inductance. This 
is much more severe for EFI ' s than 
for EBW ' s . EBW ' s can be reliably 
fired over 100 feet of twin lead or 
300 feet of coax at 3.5kv from a 1 
microfarad capacitor. To fire EFI ' s 
over 10 feet, generally requires a 
flat cable, or some other method of 
obtaining the low inductance required 
for a fast rise time* 

REFERENCES 

(1) Luis W. Alvarez, "Alvarez: Adven- 



- 122 - 



tures of a Physicist," Basic Books, 
Inc., New York, 1987, pp. 132-135. 

( 2 ) "Exploding Wires , " Vol . 1-4 , 
edited by W . G Chace and H . K . Moore , 
Plenum Press, New York, 1959 - 1968. 

( 3 ) T.J. Tucker , " Exploding Wire 
Detonators : The Burst Current Criter- 
ia of Detonator Performance," in 
Exploding Wires, Vol. 3, edited by 
W . G . Chase and H . K . Moore , Plenum 
Press, New York, 1964, p. 175. 

(4) "History," RISI Technical Top- 
ics, San Ramon, CA. , Issue 05-93. 

(5) R.S. Lee and R.E Lee, "Electro- 
static Discharge Effects on EBW 
Detonators," UCRL-ID-105644, Lawrence 
Livermore National Lab, August 1991. 

(6) Letter Report, R.H. Joppa, Los 
Alamos National Laboratory to Eglin 
Air Force Base, July 10, 1973. 

( 7 ) Carl F . Austin and Carl C . 
Halsey, "Safety and Durability Tests 
of the Fireline Explosive Cord," TS 
74-47 

(8) "Secondary Explosive Initiators & 
Accessories, " RISI Product Catalog, 
San Ramon, CA. , April 1992. 

(9) H.A. Golopol, et al., "A New 
Booster Explosive, LX-15, " UCRL- 
52175, Lawrence Livermore National 
Laboratory, March 1977. 

(10) Richard M. Joppa, et al, "Re- 
sponse of Airborne Electroexplosive 
Devices to Electromagnetic 
Radiation, " Technical Report ASD-TR- 
73-10, Los Alamos Scientific Labora- 
tory, February 1973. 

(11) "PM-25 and PM-100," RISI Data 
Sheet, San Ramon, CA. , 1994. 



- 123 - 



LOW COST, COMBINED RADIO FREQUENCY AND ELECTROSTATIC PROTECTION 

FOR ELECTROEXPLOSIVE DEVICES 

Robert L.Dow 
President, Attenuation Technology Incorporated 

Attenuation Technology Incorporated 

9674 Charles Street 

La Plata, Maryland 20646 



fie 



Abstract : Attenuation Tech- 
nology Inc. (ATI) has devel- 
oped a series of ferrite 
attenuators for protecting 
electroexplosive devices 
(EEDs) from inadvertent actua- 
tion due to RF exposure . 
ATI ' s first attenuator was 
fabricated using the MN 67 
ferrite formulation. That at- 
tenuator protected EEDs from 
both pin-to-pin and pin-to- 
case RF exposure . Those at- 
tenuators passed MIL STD 1385B 
testing when used in electric 
blasting caps (EBC) , electric 
squibs, and firing line fil- 
ters made for the US Navy. 

An improved attenuator, fabri- 
cated using ferrite formula- 
tion MN 68™, protects EEDs 
from both RF and electrostatic 
inadvertent actuation . The 
pin-to-pin and pin-to-case 
combination protection previ- 
ously demonstrated with MN 67 
attenuators was maintained for 
the RF and extended to the 
electrostatic protection area. 
In-house testing indicates 
that the EED protection can be 
extended to near-by lightning 
protection using these new at- 
tenuators. 

Franklin Research Center inde- 
pendently confirmed the in- 
creased protection provided by 
MN 68™ Ferrite Devices using 
the Mk 11 Mod EBC which uses 
a conventional bridgewire ini- 
tiator. A new R&D MN 68™ 



Ferrite Device combined with a 
Semiconducting Bridge (SCB) 
passed MIL STD 1385B testing 
for the first time. ATI is 
continuing to extend the pro- 
tection technology to other 
EED initiator designs and uses 
other than EEDs. 

ATI has issued USA Patents on 
the MN 67 and MN 68™ protec- 
tion technologies and for many 
individual applications. USA 
patent applications are pend- 
ing on SCB protection and 
other newer EED applications . 
US and overseas patent appli- 
cations are pending on the ex- 
tended coverage. 

ATI developed the technology 
for supplying ATI Certified MN 
68 Ferrite Devices for each 
application that has passed 
the required qualification 
testing . ATI has the Trade 
Mark on MN 68™. The Certifi- 
cation Mark application is 
pending at the US Patent and 
Trademark Office. Production 
tooling is available to manu- 
facture the .25 caliber MN 
68™ Ferrite Devices. 



Introduction : Attenuation 
Technology Inc. (ATI) has de- 
veloped a series of ferrite 
protection devices using a new 
and different basic technical 
approach. That approach is to 
modify the basic MN 68™ Fer- 
rite Formulation in order to 



- 125 - 



\vi. 



.,LY LiL- 



get the desired combined RF 
and electrostatic (ES) protec- 
tion characteristics in the 
final ferrite protection de- 
vice , and then to place that 
ferrite device inside the 
electroexplosive devices 
( EEDs ) in direct contact with 
and electrically grounded to 
the conductive case. 

Background : Prior to our new 
efforts in this area , prior 
art EED RF protection applica- 
tions used ferrite devices 
with Curie Temperatures as low 
as 150°C. The Curie Tempera- 
ture of these prior art de- 
vices was exceeded within a 
few seconds when the EED was 
exposed to RF energy levels 
called out in MIL STD 1385. 
The early prior art ferrite 
devices also had cutoff fre- 
quencies above 3 megahertz, 
which made them unacceptable 
when MIL STD 1385 required RF 
protection at one megahertz. 

These early failures gave 
generic ferrite protection de- 
vices a bad reputation that 
they still have difficulty 
overcoming. Many people still 
automatically revert to think- 
ing of these early failures 
when the sub j ect of EED pro- 
tection using any ferrite de- 
vices is discussed. As a con- 
sequence, they are still ex- 
cluded as alternative EED 
protection options even though 
the technology has changed 
significantly. As far as ATI 
can determine , we are the 
first company to specially 
formulate a specific ferrite 
formulation, MN 68™, for EED 
protection and then make spe- 
cific formulation modifica- 
tions for individual EED ap- 
plications. 



New Technology : The first 
ferrite formulation character- 
istic that was changed was the 
Curie Temperature. This phys- 
ical property characteristic 
is important because, when the 
lossy ferrite device is ex- 
posed to RF energy, it con- 
verts that RF energy to heat. 
If that heat cannot be removed 
effectively , the ferrite de- 
vice will increase in tempera- 
ture. When the ferrite device 
reaches its Curie Temperature, 
it stops converting the RF en- 
ergy to heat. If the ferrite 
device reaches its Curie Tem- 
perature, it can not protect 
the EED. The RF attenuation 
property is reversible in that 
once the ferrite device cools 
below its Curie Temperature, 
it resumes attenuating RF en- 
ergy until it repeats the tem- 
perature cycle. 

ATI has consistently increased 
the Curie Temperatures of its 
lossy , soft ferrite formula- 
tion with 250 to 280°C modifi- 
cations currently available 
for manufacture. ATI is work- 
ing toward a goal of at least 
approaching 4 00°C on a second 
ferrite formulation series . 
Progress has been slow on this 
new ferrite formulation, since 
ATI appears to be the only 
customer in the USA with in- 
terest in the very high Curie 
Temperature , lossy f err ites . 
There is some interest devel- 
oping in Europe in using these 
very high Curie Temperature 
formulations. 

The second lossy ferrite for- 
mulation characteristic that 
was changed was to provide a 
controlled DC resistance in 
the ferrite device. Most of 
the previous art EED ferrite 
protective device applications 
used lossy f err ites that were 



- 126 - 



not DC conductive . Further , 
most of these prior art lossy 
ferrite devices were potted in 
place in the EEDs with noncon- 
ductive adhesives. We made 
the conscious effort to go in 
the opposite direction and 
provide lossy, high Curie Tem- 
perature ferrite device with a 
controlled DC resistance 
within an acceptable range. 

The third lossy ferrite formu- 
lation physical property that 
is included in all our ferrite 
formulations is cut off fre- 
quencies below one megahertz. 
MN 68™ Ferrite Devices suc- 
cessfully attenuate RF energy 
at 10 kilohertz. 

Our lossy ferrite devices have 
a sufficiently low DC resis- 
tance to equalize the electro- 
static energy that can build 
up between the firing leads of 
an EED (pin-to-pin) and/or be- 
tween the firing leads and the 
conductive case of an EED 
(pin-to-case) . Our lossy fer- 
rite devices have sufficiently 
high DC resistance so that 
they do not act as a DC shunt 
for the EED's DC firing pulse. 
Each EED application must be 
tailored to work within those 
DC limits. ATI has developed 
the technology to provide this 
acceptable range for each EED 
application. 

Improved EED Design ; Once the 
three ferrite device require- 
ments were met, the design of 
the RF protective device and 
assembly methods used for se- 
curing it in the EED could be 
greatly simplified. Noncon- 
ductive potting materials were 
no longer required during EED 
assembly. The electrical in- 
sulation previously used on 
the firing leads passing 
through the ferrite device was 



eliminated. One lossy ferrite 
device could provide both RF 
and ES protection for the EED 
for both pin-to-pin and pin- 
to-case RF and ES energy expo- 
sure modes. 

It was also determined that 
the ferrite device could be 
inserted into EEDs, such as 
electric blasting caps ( EBC) 
thin conductive case , without 
excessive breakage . It was 
further determined that a good 
electrical ground could be 
achieved between the lossy 
ferrite device and the conduc- 
tive case using just the in- 
sertion assembly method. All 
of this work was done with 
ferrite devices that were 
right circular cylinders. New 
manufacturing methods have re- 
cently been developed, so that 
a chamfer can be molded into 
the finished ferrite device to 
make assembly of other EED de- 
signs easier. 

Benefits of Design ; The first 
synergistic effect of this as- 
sembly procedure was that once 
all of the electrically insu- 
lating potting material was 
removed from the EED design, 
the heat conduction path 
available to cool the RF at- 
tenuating ferrite device were 
great ly improved . Thus , sub- 
sequent EED designs that con- 
tained the higher Curie Tem- 
perature lossy ferrite de- 
vices , actually stabilized at 
lower temperatures when ex- 
posed to comparable amounts of 
RF energy than prior art EED 
designs. 

This observation was at- 
tributed to an improved heat 
removal path directly to the 
EED's conductive case, provid- 
ing better cooling of the at- 
tenuating, lossy ferrite 



- 127 - 



device. Thus, the safety fac- 
tor of these new EED designs 
were improved by two mecha- 
nisms instead of the original 
approach of simply using the 
higher Curie Temperature , 
lossy ferrite devices. 

The second synergistic effect 
was that tieing the conductive 
EED case to the firing leads 
through the controlled circuit 
ferrite allowed large amounts 
of ES energy to be safely dis- 
sipated without firing the 
EED . Laboratory tests of the 
ATI ferrite devices showed 
that repeated exposures of the 
wound ferrite devices with up 
to 12 Joules of ES energy did 
not destroy the ferrite device 
or change its RF attenuation 
properties. Most other compo- 
nents in the EED when exposed 
to that level of ES energy 
only once, were completely 
disintegrated , destroyed , or 
failed to the duded mode. 

The final discovery was that 
the level of RF protection 
could be changed by selection 
of the winding pattern used in 
the ferrite device. One and 
one half turns of additive 
choke windings on each firing 
lead was necessary for the EBC 
to pass MIL STD 1385B environ- 
ments , but other winding pat- 
terns can be used for commer- 
cial applications where the RF 
exposure hazards are lower. 

It was independently deter- 
mined that the improved fer- 
rite devices did not signifi- 
cantly attenuate the DC firing 
pulse , even when they were 
used to protect the semicon- 
ducting bridge (SCB) igniters 
that use microsecond DC firing 
pulses . One design of the Mk 
66 Igniter , containing a sin- 
gle ATI high Curie Temperature 



lossy ferrite choke , has been 
reported as passing MIL STD 
1385B RF testing as well as 
the electrostatic testing. 

Production Status : ATI is 

currently working with three 
USA ferrite device manufactur- 
ers to produce these lossy , 
high Curie Temperature, con- 
trolled DC resistance, ferrite 
devices using high volume, low 
cost production techniques . 
So far, the largest production 
lot has only been 25,000 bare 
ferrite chokes . This quan- 
tity, while very small by fer- 
rite manufacturer's standards, 
was produced without signifi- 
cant production problems. ATI 
is also working with an over- 
seas source for potential ap- 
plications in Europe. 

Certification Approach ; Some 
of the ferrite manufacturers 
are offering their versions of 
high Curie Temperature, lossy 
ferrite devices that are pur- 
ported to be as effective as 
our MN 68™ Ferrite Devices. 
We have been awarded the USA 
Trademark on MN 68™ to dis- 
tinguish our ferrite devices , 
certified by ATI , from those 
produced with similar formula- 
tions but not tested as com- 
prehensively. ATI has a Cer- 
tification Mark pending at the 
US Patent and Trademark Of- 
fice. 

Since this technology niche 
has not been investigated be- 
fore , ATI was forced to de- 
velop the techniques and 
measurement equipment on its 
own to measure the performance 
and certify the effectiveness 
of these ferrite devices. ATI 
has developed these to the 
point where, once a specific 
ferrite device has been quali- 
fied for a specific EED appli- 



- 128 - 



cation , subsequent production 
lots can be certified to the 
levels required. 

ATI is maintaining certified 
ferrite device samples from 
each EED successfully quali- 
fied. These samples are main- 
tained to assure that new lots 
of the ferrite devices pro- 
duced in future years can be 
directly compared to the orig- 
inal and certified equivalent 
in performance to the baseline 
sample. If the project spon- 
sor wants improved performance 
(based on how the technology 
has progressed in the mean- 
time) or the projects safety 
requirements have increased in 
the interim, an improved ver- 
sion ferrite device can also 
be manufactured and made 
available. It now appears 
that it may be possible to 
produce the improved perfor- 
mance ferrite devices without 
any modifications to the pro- 
duction tooling. 

US Patents : Since all of this 
work has been funded solely by 
ATI , patent protection is the 
main form of intellectual 
property rights protection . 
ATI's issued patents are: 

1. US 5.036,768 Basic Patent 
on MN 68™ Applications 

2. US 5,243,911 Near-by Light- 
ning Protection for EED 

The main, generic approach, 
patent application revealing 
the principles of EED protec- 
tion regardless of the con- 
trolled property ferrite for- 
mulation used or the winding 
pattern employed is expected 
to be issued shortly. ATI has 
other patent applications 
pending in the areas of EED 



protection and ferrite device 
certification areas. 

During the EED protection 
technology evolution, ATI de- 
termined that the technology 
can be modified to protect 
other devices as well . The 
first medical protection de- 
vice patent US 5,197,468 has 
been issued. Other patent ap- 
plications are pending cover- 
ing many classes of equipment. 

Conclusion : Since this is a 
completely different approach 
to protecting EEDs from both 
RF and ES inadvertent ignition 
with a single device , ATI is 
prepared to discuss any poten- 
tial applications of this new 
technology area with anyone 
interested . Each solution 
must be tailored for each ap- 
plication. The technology ap- 
pears to have progressed suf- 
ficiently so that can be done. 
Please contact us if you are 
interested in considering this 
new approach. 



- 129 - 



SI z-z% 



Unclassified 



Distribution Unlimited 



SAND94-0246C 



UNIQUE PASSIVE DIAGNOSTIC FOR SLAPPER DETONATORS 

William P. Brigham 
Explosive Projects and Diagnostics Department 

John J. Schwartz 
Stockpile Evaluation Department 

Sandia National Laboratory 
Albuquerque, New Mexico 



ABSTRACT 

The objective of this study was to find a material and configuration that could 
reliably detect the proper functioning of a slapper (non- explosive) detonator. 
Because of the small size of the slapper geometry (on the order of a 15 mils), 
most diagnostic techniques are not suitable . This program has the additional 
requirement that the device would be used on a centrifuge so that it could not 
use any electrical power or output signals. This required that the diagnostic 
be completely passive. 

The paper describes the three facets of the development effort: complete 
characterization of the slapper using VISAR measurements, selection of the 
diagnostic material and configuration, and testing of the prototype designs. 
The VISAR testing required the use of a special optical probe to allow the laser 
light to reach both bridges of the dual-slapper detonator. Results are given in 
the form of flyer velocity as a function of the initiating charge voltage level. 
The selected diagnostic design functions in a manner similar to a dent block 
except that the impact of the Kapton disk from a properly- functioning slapper 
causes a fracture pattern. A quick visual inspection is all that is needed to 
determine if the flyer velocity exceeded the threshold value. Sub -threshold 
velocities produce a substantially different appearance. 



Introduction 

Slapper-detonators are used in current 
weapon systems because of their fast 
function time , small j itter , and 
relatively low energy requirements 
compared to other types of detonators. 
As part of the Sandia National 
Laboratories (SNL) evaluation program, 
detonators are typically tested in the 

This work was supported by the United 
States Department of Energy under 
Contract DE-ACO4-94AL85000. 



laboratory under realistic physical 
and environmental conditions . The 
unique operation and design of the 
slapper requires sophisticated 
diagnostics like VISAR or closure 
switches to evaluate performance 
parameters. The complexity of these 
techniques makes their use difficult 
in the evaluation lab. This paper 
discusses a device developed to 
provide a simple evaluation of slapper 
function within the constraints 



- 131 - 



p/ 



\U: ..J. 



|J0 



j;:jV p! r-*i£ 



imposed by the laboratory environment. 
It consists of a small glass target 
that provides a unique visual record 
when impacted by the flyer from the 
slapper . 



Slapper Detonator 

The SNL slapper detonator is shown in 
Figure 1. It consists of a central 
"bullseye" connector for attachment to 
the firing set with the flat copper 
cables extending in opposite 
directions to form a single loop. The 
upper layer of thin copper narrows at 
each end to form a "bridge" that 
causes the current density to increase 
significantly. The current through 
each bridge is strong enough to drive 
the copper into vapor. The pressure 
of the gas causes the Kapton to shear 
against the sapphire "barrel" forming 
a rapidly-accelerating flyer with a 
diameter of about 0.015". 

Flyer velocity is a function of the 
initial firing set charge voltage and 
the resulting current . Typical 
terminal velocity is on the order of 
3-4 mm//is, although other designs are 
capable of nearly 6 mm/^s . It is 
assumed that the flyer begins to come 
apart soon after it leaves the barrel, 
but good data have been obtained for 
distances up to 1 mm. The most common 
practice is to place the explosive of 
the next assembly in contact with the 
barrel. 



Slapper Characterization 

Prior to the development of the 
passive diagnostic , it was necessary 
to determine the performance 
(velocity-time history) of the slapper 
detonator over a range of initial 
firing set voltage. The best tool for 
this measurement is VISAR (Velocity 
Interferometer System for Any 
Reflector) that uses Doppler- shifted 
laser light from the slapper surface 
to infer the velocity. Because this 



slapper is a dual -bridge functioned 
from a common firing set, it was 
desired to measure both flyers 
simultaneously. Most VISAR' s are 
"dual -leg" to provide redundant 
measurement of a single device using 
two interferometers with different 
experimental constants . A unique 
optical probe was therefore developed 
to convert the existing equipment into 
a dual system to make the two separate 
measurements . 

The probe allows the light from a 
single laser to be split into two 
equal beams that are then fiber - 
coupled to the target locations . A 
simplified schematic of the probe is 
shown in Figure 2 to demonstrate the 
path of the light. The input fiber 
connects to optics that allow the 
light to be focused onto the target 
surface . The target surface causes 
the light to be scattered diffusely 
where it is collected by other optics 
within the probe. The light is then 
collimated and focused onto the return 
fiber located at the rear of the 
probe. In this manner the image from 
each slapper can be routed to a 
different VISAR leg. 

A unique feature of the optic probe is 
the incorporation of a small camera to 
give a TV image of the target surface. 
The selective coating on the angled 
mirror causes a portion of the 
returned light to be transmitted to 
the camera element. The TV image is 
then present on a monitor within the 
test area. This feature is essential 
when testing slappers with a small 
active area because it allows the 
operator to precisely align the input 
laser light onto the Kapton element in 
the center of the barrel. Placing 
each end of the slapper on a separate 
translation stage allows adjustment of 
the position just prior to the test. 

Additional diagnostics include 
detection of the charge voltage and 
current waveforms from the firing set. 
Each of these along with the VISAR 



- 132 - 



data were recorded on Tektronix DSA 
602 transient digitizers. Data 
reduction was performed following each 
test and stored on computer disk. 

Figure 3 shows VISAR data from an 
experiment where the initial charge 
voltage was 2.6 kV. The data are in 
the form of flyer velocity and 
displacement versus time where the 
distance is obtained by numerical 
integration of the velocity- time 
record. The behavior is typical of 
slapper detonators in that the 
acceleration of the flyer is less 
abrupt than a conventional 
explosively-driven plate, although the 
final velocity at the exit of the 
barrel is over 3 mm//is . Note that the 
record continues for a distance 
approaching 0.75 mm. At the higher 
initial charge voltage levels, breakup 
of the flyer is assumed as manifested 
by increasingly noisey signal quality. 

The data of Figure 3 can be cross - 
plotted to give velocity as a function 
of displacement as shown in Figure 4 . 
This is particularly useful in 
assessing the velocity as a given 
distance , such as the exit of the 
barrel or the location of the next 
assembly. 

The firing set for all tests was a Hi- 
Voltage Components, Inc. model CDU2045 
with a 0.2 /ifd, 5-kV capacitor. 
Initial charge voltage was varied from 
1.6 kV to 3.0 kV. Two tests were done 
at each voltage level to provide 
redundant data. 

Results for the characterization test 
matrix are given in Figure 5 in the 
form of flyer velocity as a function 
of firing-set charge voltage. The 
numerical results from the same tests 
are shown in Table 1. The velocity is 
obtained at a propagation distance of 
0.5 mm, or just beyond the exit of the 
barrel. The four data points at each 
voltage represent the measurement of 
each side of the slapper, with two 
tests at each level . In all cases , 



the designation "A" refers to the 
bridge that first receives the firing 
pulse , based on the direction of the 
current flow. The "B" suffix then 
denotes the opposite bridge. 

The results given in Figure 5 indicate 
that the behavior of the detonator is 
significantly more erratic at voltage 
levels below 2 kV. It is thought that 
at the lower voltage, minor tolerance 
differences in the construction of the 
bridges have a more pronounced effect 
on the behavior. At the higher 
voltage levels , sufficient energy is 
available to overcome these 
differences and the velocity achieved 
by the two bridges is more consistent. 
For both tests at 1.6 kV, the "B" side 
of the unit failed to cause 
acceleration of the Kapton to a 
velocity sufficient to be measured by 
the VISAR, which has a lower detection 
limit of about . 2 mm//is . Inspection 
of the bridge following the shot 
indicated that the copper had failed 
in the region of the bridge. 

Results in the voltage range above 2.0 
kV show a correspondence between the 
applied voltage and the resulting 
velocity. The greatest velocity, on 
the order of 4 mm//is, was achieved at 
the highest voltage level. Some 
scatter is present above 3 . 5 mm//is 
that is probably caused by breakup of 
the Kapton flyer. There does not 
appear to be a discernible trend to 
establish that side "A" or "B" is 
consistently higher in velocity at a 
given voltage. 



Passive Diagnostic 

As mentioned previously , the purpose 
of this work was to develop a passive 
diagnostic for the slapper used during 
evaluation testing. The specific 
requirements were that the device 
would not require any external 
communication (input power or signal 
cables) and that the visual indication 
could be easily detectable. Previous 



- 133 - 



experience had shown that brittle -like 
fracture could be introduced into some 
plastic or ceramic materials using 
metallic flyers from hot-wire 
detonators. This observation 

suggested that a similar arrangement 
could be used for the Kapton flyer 
from the slapper. 

Initial screening tests looked at a 
variety of material types, sizes, and 
thicknesses . The presumed 50% 
probability threshold velocity for 
detection corresponded to an initial 
voltage of about 2.1 kV, although 
tests were done at higher and lower 
levels to ascertain the sensitivity of 
each configuration. Table 2 contains 
the results from the initial screening 
tests . Some of the tests used a 
single -bridge and are shown as one 
configuration only, while the 
remainder are dual -bridge that may 
have a different device on each end. 

The screening matrix demonstrated a 
number of viable candidates for 
selection based on the appearance of 
obvious cracks above the threshold 
voltage level. The best material was 
sapphire, but was rejected because of 
prohibitive expense. The 0.036" thick 
microscope slide also showed excellent 
performance, but a source for this 
material could not be located. The 
next alternative was to obtain 
standard fused silica material in 
thicknesses in the range of 0.020" to 
0.030". Although this is thinner than 
typically produced, several vendors 
were found that could grind 1" 
diameter disks to virtually any value. 
Disks were purchased in . 020" and 
0.025" thicknesses of BK-7, a standard 
fused silica composition made for the 
optics industry. A total of 16 tests 
were conducted, six using the 0.020" 
thickness and the remainder with the 
0.025" pieces. Table 3 shows the 
results of the BK-7 tests. 

The 0.020" thick BK-7 samples from 
tests at 2.2 and 2.4 kV demonstrated a 
large number of cracks at both voltage 



levels, while the 0.025" samples only 
showed cracks at the higher charge 
voltage. The behavior described above 
was relatively consistent throughout 
the remaining tests . One additional 
modification was made for the last 
nine tests of the shot series . A 
simple fixture of aluminum was made to 
hold the glass against the slapper, 
using a 1/16" thick 0-ring to provide 
a mild compressive load. The other 
side of the slapper continued to use 
the glass glued to the sapphire 
barrel. This was done to determine if 
a fixture would have any effect on the 
crack initiation and propagation. 

The results with the fixture suggest 
that i t may have induce a s 1 i ght 
increase in the crack sensitivity of 
the glass . This is not consistently 
true for all the shots, but in no case 
did the unfixtured end show greater 
damage than the corresponding fixtured 
side. 



Conclusion 

As part of the development of a 
passive diagnostic for surveillance 
testing of a slapper detonator, a 
unique optical probe was constructed 
that allows simultaneous VISAR 
measurement of both flyers . The 
characterization of the slapper was 
done to establish the correlation 
between applied firing voltage and the 
resulting flyer velocity. At voltages 
below 2 kV, significant differences 
were seen between the two sides of the 
unit and from one unit to another . 
This was attributed to minor 
differences in the construction of 
each device. 

A variety of materials were tested to 
determine which would provide a 
discernible crack pattern at flyer 
velocities above a predetermined 
threshold value . Standard fused 
silica was selected based on 
performance, availability, and cost. 
For this slapper, a thickness of 



- 134 - 



0.020" to 0.025" and a diameter of 1" 
provides acceptable performance . 
Efforts are continuing using an 
intermediate thickness and fixturing 
to improve the device's behavior for 
the intended application. 



Table 1 
Slapper Flyer Velocity Results 



Table 3 
BK-7 Test Results 



Firing Set 
Voltage (kv) 


Velocity 

Bridge A 

(mm/ms) 


at 0.5 mm 

Bridge B 

(ram/fis) 


1.6 
ti 


2.37 
2.20 


- 


1.7 

tl 


2.90 
2.52 


1.72 
2.52 


1.8 

II 


2.55 
3.06 


2.70 
2.50 


1.9 

II 


3.15 
2.85 


2.66 
2.76 


2.0 

II 


2.88 
3.05 


2.93 
3.08 


2.1 

II 


3.12 
3.26 


3.15 
2.98 


2.2 

ll 


3.36 
3.27 


3.52 
3.33 


2.4 

II 


3.58 
3.52 


3.52 
3.44 


2.6 

II 


3.68 
3.50 


3.88 
3.60 


2.8 

II 


3.78 
3.68 


3.96 
3.84 


3.0 
ii 


4.05 
4.10 


4.14 
3.85 



Charge 

Voltage 

(kV) 


Sample 

Thickness 

(in.) 


Number of 
Cracks 
(A) (B) 


2.2 


0.020 








2.4 


tf 


4 


6 


2.6 


II 


5 


6 


2.2 


0.020 


6 


many 


2.0 


« 


4 


2 


1.9 


II 


3 


5 


1.8 


■1 








2.2 


0.025 


6 


8* 


2.0 


It 





0* 


2.1 


II 





0* 


2.2 


II 





o* 


2.3 


II 





o* 


2.4 


■■ 


3 


9* 


2.3 


it 


2 


0* 


2.0 


0.020 


5 


0* 


2.2 


■i 


4 


0* 



* indicates fixture was used 
on Side B 



- 135 - 



Table 2 
Results From Initial Screening Tests 



Shot 
Number 


Charge 
Voltage 
(kv) 


Sample Configuration 


Results 


1 


1.8 


0.03" thk PMMA 


surface damage 


2 


2.8 


tl 


surface damage 


3 


2.8 


0.036" microscope 
slide 1/2" x 3/4" 


numerous cracks 


4 


1.8 


tt 


4 cracks from 
contact point 


5 


2.0 


.006" cover glass 
22 mm square 


4 cracks 






.065" quartz glass 
1" square 


no damage 


6 


2.8 


.006" cover glass 
22 mm square 


>10 cracks 






.065" quartz glass 
1" square 


3 cracks 


7 


2.0 


.039" microscope 
slide 1" square 


damage in center 
no cracks to edge 






silicon substrate 
.016" x -1" square 


6 cracks 


8 


2.8 


.039" microscope 
slide 1" square 


damage in center 
no cracks to edge 






silicon substrate 
.016" x -1" square 


many cracks , 
center missing 


9 


2.8 


plate glass 1/8" 


no damage 






.036" microscope 
slide 1/2" x 3/4" 


6 cracks 






(continued) 





- 136 - 



Table 2 (Continued) 



10 


2 


.8 


.039" microscope 
slide 1" square 

.036 microscope 
slide 1" square 


minor damage 
4 cracks 


11 


2 


.0 


.062" plain glass 
-1" square 

.020" x 3/4" dia 
sapphire disk 


no damage 
no damage 


12 


2 


.8 


.062" plain glass 
-1" square 

.020" x 3/4" dia 
sapphire disk 


no damage 

3 cracks , rear 
spall 


13 


2 


.4 


.065" quartz -1" 
square 

.020" x 3/4" dia 
sapphire disk 


no damage 

1 diagonal crack 
edge to edge 


14 


2 


.2 


.038" quartz ~1" 
square 

.020" x 3/4" dia 
sapphire disk 


1 crack, edge to 
edge 

incipient fracture 
no complete cracks 


15 


2 


2 


.018" quartz, -1" 
square 

.038" quartz ~1" 
square 


several cracks, 
not from center 

incipient cracks 


16 


1 


95 


.019" Dynasil, odd 
shape 

.019" Dynasil, odd 
shape 


6 cracks 
5 cracks 


17 


1 


95 


.034" Dynasil, odd 
shape 

.034" Dynasil, odd 
shape 


no damage 
no damage 



- 137 - 



V>4 
OO 



Isolator 

2 ml Kapton XT 
ImlPyrafcut 

Saphire Barrel 

IB ml 10 

Adhesive 

1 mlPyrafcix 

Bridge/Flyer 

1 ml Kapton Ryar 
175 Mftcroinch Cu 

Adhesive 

2mlPyraiux 



Top Cover Coat 

2 ml Kapton 
2miPyrafci* 




Circuit "A" 

1 oi. Cu 
Epoxy 
2ml Kapton 



Inside Adhesive 



1 ml PynAix 
1 ml Kapton 
1 ml Pyrahix 

Circuit "BC" 

1 oi. Cu 
Epoay 
2ml Kapton 



Bottom Cover Coat 



2 ml Pyrafci* 
2 ml Kapton 



SLAPPER DETONATOR 



GB 



Sanda 
national 

laboratories 



figure 1 



IMAGING FIBER OPTIC COUPLED SENSOR 



INPUT FIBER OPTIC FROM LASER 



TARGET 







RETURN FIBER 
OPTIC TO 
VISAR 



SECTIONAL VIEW OF SENSOR. LASER LIGHT FROM SOURCE ILLUMINATES TARGET, REFLECTED 
LIGHT IS COLLECTED AND FOCUSED INTO RETURN FIBER OPTIC FOR ANALYSIS. PAT. PEND. 



figure 2 



lELOCITV/DISPLflCEMENT US TIME LEG "A" 2.6kU Date : 81/2B/94 



0.15 



CD 
I 



a 
to 
W 

3 
\ 

£ 
S 



a 
o 

Of 

3 



D.H5 




W 

■D 

o 

3 



0.15 



150 

TIME <nS) 



figure 3 



JELOCITY US DISPLACEMENT LEG "A" 2.6 kU CHARGE Date: 81/28/94 



I 

t— » 

-pr- 
I— » 

I 



O 
to 
W 

\ 

£ 



3* 


o 

Op 



35 












3 












2.5 












g 












15 












| 












5 












D 













0.15 



0.3 D.H5 

Displacement (mm) 



D.E 



0.15 



figure 4 



SLAPPER DETONATOR 

VELOCITY @ .5mm 



4.5 






(0 

"e 

E, 

>- 

H 
O 

o 

_l 
LU 
> 



4- 



3.5- 



2.5- 



2- 



1.5 



1.4 1.6 




1.8 2 2.2 2.4 2.6 2.8 
CHARGE VOLTAGE (KILOVOLTS) 



SHOT 1 A 

-~EEH- 

SHOT 1 B 
SHOT 2A 
SHOT 2B 



3.2 



figure 5 



APPLYING ANALOG INTEGRATED CIRCUITS 
FOR HERO PROTECTION 

By: 

Kenneth E. Willis 

Quantic Industries/ Inc 

990 Commercial Street 

San Carlos, CA 94070 

(415) 637-3074 

Thomas J. Blachowski 

Naval Surface Warfare Center 

Indian Head, MD 

(301) 743-4876 






INTRODUCTION 

One of the most efficient methods 
for protecting electro-explosive de- 
vices (EEDs) from HERO and ESD is 
to shield the EED in a conducting 
shell (Faraday cage). Electrical en- 
ergy is transferred to the bridge by 
means of a magnetic coupling 
which passes through a portion of 
the conducting shell that is made 
from a magnetically permeable but 
electrically conducting material. 
This technique was perfected by ML 
Aviation, a U.K. company, in the 
early 80 ! s, and was called a Radio 
Frequency Attenuation Connector 
(RFAC). It is now in wide use in 
the U.K. Previously, the disadvan- 
tage of RFAC over more conven- 
tional methods was its relatively 
high cost, largely driven by a thick 
film hybrid circuit used to switch 
the primary of the transformer. 

Recently, through a licensing 
agreement, this technology has 
been transferred to the U.S. and 
significant cost reductions and per- 
formance improvements have 
been achieved by the introduction 
of analog integrated circuits. 



An integrated circuit performs the 
following functions: 1) Chops the 
DC input to a signal suitable for 
driving the primary of the trans- 
former, 2) Verifies the input volt- 
age is above a threshold, 3) Verifies 
the input voltage is valid for a pre- 
set time before enabling the device, 
4) Provides thermal protection of 
the circuit, and 5) Provides an ex- 
ternal input for independent logic 
level enabling of the power transfer 
mechanism. This paper describes 
the new RFAC product and its ap- 
plications. 

BACKGROUND 

Electro-explosive devices (EEDs) 
must, in many applications, be pro- 
tected against unintended initia- 
tion by Electromagnetic Radiation 
(EMR) or Electrostatic Discharge 
(ESD). 

A variety of methods are in use to 
mitigate these problems; they in- 
clude: 

1) Low resistance bridgewires 
which can dissipate some elec- 



- 143 - 



trical energy before heating to 
ignition temperature. 

2) Filters consisting of capacitive 
or inductive elements which 
can absorb or reflect RF energy. 

3) Shielding to create a Faraday 
cage around power source, con- 
ductors and the EED. 

4) Voltage spike dissipation or 
"clamping" functions. 

5) Switches or relays to discon- 
nect/short the EED pins until 
ready for function. 

All of these solutions suffer from 
one or more of the following defi- 
ciencies: 

1) Not completely effective. 

2) Impose undesirable constraints 
on the system, e.g., weight, 
power, and envelope. 

3) Adds cost. 

Depending on the system require- 
ments, these deficiencies can be 
more or less annoying. 

The ideal solution is to simply sur- 
round the EED with a Faraday cage 
- cheap and 100% effective. This so- 
lution leaves only one problem: 
how do you get the firing energy 
into the bridgewire when it is sup- 
posed to function. A practical im- 
plementation of this concept is 
shown in Figure 1. 

The Faraday closure surrounds the 
EED and the secondary of the trans 




WTVT MLftt CAPAOTM 



PRACTICAL IMPLEMENTATION 

Figure 1 

former. The material between the 
transformer cores is a conducting, 
but magnetically permeable, alloy. 
The secondary of a narrow band- 
pass transformer inside the Faraday 
closure generates the firing current. 
The conducting copper-nickel alloy 
maintains the integrity of the Fara- 
day cage. The primary of the trans- 
former is driven with an AC signal 
at the mid-frequency of the pass 
band, generated by the DC-AC in- 
verter. 

This concept, of course, is not new. 
The physics has been around a long 
time. The application to EED pro- 
tection, as near as I could trace its 
origins, was proposed by Wing 
Commander Reginald Gray of the 
Royal Air Force in 1957. In the mid 
to late 1970's, Mr. Raymond Sell- 
wood of ML Aviation, a U.K. de- 
fense company, adapted this con- 
cept to a practical device called the 
Radio Frequency Attenuating Con- 
nector (RFAC). Mr. Sellwood was 
then granted several patents for 
these designs, including a U.S. 
Patent in 1979. 

The RFAC has been successfully 
deployed on a number of U.K. 
weapon systems, and has per- 



- 144 - 



formed flawlessly- These systems 
include: 

• Chevaline SLBM 

• Airfield Attack Dispenser 
(Tornado) 

• Torpedoes (Spearfish and 
Stingray) 

• Stores Ejection Systems 

• ASDIC (Cormorant) 

In 1989, Mr. Tom Blachowski, this 
paper's co-author, completed test- 
ing of an RFAC equipped impulse 
cartridge for a Naval Surface War- 
fare Center (Indian Head) applica- 
tion. 

EVALUATION TESTING 

The Indian Head Division, Naval 
Surface Warfare Center (NSWC) 
completed an evaluation of the in- 
ductive coupling technology for 
electrically actuated cartridges and 
cartridge actuated devices (CADs). 
This effort was performed as part of 
the Naval Air Systems Command 
Foreign Weapons Evaluation 
(FWE)/ NATO Comparative Test 
Program (CTP). The FWE Program 
is designed to assess the applicabil- 
ity for foreign-developed, off-the- 
shelf technology for procurement 
and implementation in the U.S. 
Fleet. The FWE goal is production 
procurement offering fleet users 
enhanced performance while low- 
ering the per item cost to the pro- 
gram managers. 

The FWE effort to analyze the in- 
ductive coupling initiation tech- 
nology was structured as follows: A 
Navy standard electrically initiated 
cartridge, the Mark 23 Mod im- 
pulse cartridge, was selected to be 



modified to accept the inductive 
coupling initiation hardware. The 
MK23 Mod cartridge was previ- 
ously tested by the Navy in nu- 
merous configurations and rated as 
"susceptible" when subjected to 
Hazards of Electromagnetic Radia- 
tion to Ordnance (HERO) electrical 
field strengths. There have been 
several documented instances in 
which MK23 Mod cartridges have 
inadvertently actuated when sub- 
jected to a HERO or electromag- 
netic interference environment. 
These inadvertent actuations re- 
sulted in mission aborts, loss of es- 
sential equipment, and an in- 
creased threat to crew members and 
ground personnel during an elec- 
trical event. ML Aviation Limited 
was contracted by NSWC to install 
the inductive coupling hardware 
into the existing MK23 Mod car- 
tridge envelope maintaining the 
same form, fit, and functions as the 
Navy standard device. ML Avia- 
tion packaged the existing RFAC 
secondary transformer into the 
MK23 Mod impulse cartridge and 
designed a primary transformer 
into an electrical connector which 
must be installed in the cartridge 
firing circuit. 

These primary electrical connectors 
and inductively coupled MK23 
Mod impulse cartridges were sub- 
jected to a specialized design verifi- 
cation test series at the Indian Head 
Division, NSWC. Two phases of 
design verification test were con- 
ducted: the first phase was electrical 
requirements and analysis 
(performed in accordance with 
MIL-I-23659C, "General Design 
Specification for Electrical Car- 



- 145 - 



bridges"), and the second phase was 
the functional testing of the car- 
tridges (performed in accordance 
with MIL-D-21625F, "Design and 
Evaluation of Cartridges and Car- 
tridge Actuated Devices"). 

All of the inductively coupled 
MK23 Mod impulse cartridge tests 
were successful and the results ex- 
hibited the potential to implement 
the inductive coupling technology 
in a wide range of electrically initi- 
ated cartridges and CADs. The In- 
dian Head Division, NSWC pub- 
lished the results of this effort in 
Indian Head Technical Report 
(IHTR) 1314 dated 17 November 
1989, "Evaluation of the Inductive 
Coupling Technology Installed in a 
Standard Impulse Cartridge Mark 
23 Mod 0". 

Based on the excellent test results, 
and the Navy's desire for a U.S. 
producer for the RFAC, Quantic 
Industries and ML Aviation en- 
tered into a license agreement for 
U.S. and Canadian production and 
sales of RFAC. 

DESIGN 

A typical configuration of the 
RFAC, used in a connectorized ver- 
sion is shown here (Figure 2). The 
primary electronics module is 
housed in a 7mm diameter x 
35mm long tube. The magnetics in 
this configuration will transfer a 
minimum of five watts to the sec- 




RFAC ASSEMBLY 

Figure 2 

ondary bridgewire. The electronics 
can drive larger magnetics in a 9, 
11, or 14mm outside diameter con- 
figuration, to provide more power. 

In the original ML Aviation design, 
a self oscillating feedback circuit 
was implemented in a thick film 
hybrid. Quantic implemented sub- 
stantial cost reductions and per- 
formance improvements by replac- 
ing the original thick film hybrid 
circuit with an analog application 
specific integration circuit (ASIC). 
Relatively new technology in high 
voltage analog array ASICs made 
this economically viable. 

The electromagnetic attenuation 
performance of the RFAC is un- 
changed by the change in electron- 
ics. Sixty (60) to 80 dB attenuation is 
achieved. A typical attenuation per- 
formance is shown here in Figure 
3. 



- 146 - 



i 














I 


















j 


M 


/ 


\l\ 














lj 


rJ 


V 


In 


^ * 








< J 


/^ 










^ 


lV\ ^ 


/ 


^/ 














** "^ 




/'■- 



























o sc«o«oaoo»oaooMo«o«9cia 

RFAC PERFORMANCE 

Figure 3 

Some additional safety and per- 
formance features were added to 
the ASIC, which, incidentally, adds 
no cost to the product. These fea- 
tures include: 

1. Self contained oscillator. 

- Stable over wide tempera- 
tures and voltages 

- Simpler magnetics: original 
design used - feedback loop 
to self oscillate 

2. Programmable delay time - an 
additional guard against voltage 
transients. 

3. Input voltage threshold test. 

4. Output enable - separate logic 
level input needed to transfer 
power. 

5. Thermal cut-off. 

6. High current output - can be 
used for larger units. 

7. ESD and over voltage protection 
(note that the ordnance section 
is intrinsically SAFE from ESD. 

8. Programmable turn-off time (0- 
10 ms). 



9. Designed to be nuclear hard. 

A block diagram of the RFAC is 
shown in Figure 4: 




RFAC SIMPLIFIED ELECTRICAL 
BLOCK DIAGRAM 

Figure 4 

CURRENT STATU? 

At this time (November 1992) the 
development and engineering tests 
of the enhanced inductively cou- 
pled MK23 Mod cartridge are 
complete. The Indian Head Divi- 
sion, NSWC is conducting a quali- 
fication program to allow for pro- 
duction procurement and imple- 
mentation in a wide variety of fleet 
applications. The inductively cou- 
pled MK23 Mod cartridge has 
been renamed the CCU-119/A Im- 
pulse Cartridg e as part of this pro- 
gram. The functional test phase of 
the qualification program is again 
based upon MIL-I-23659C and MIL- 
D-21625. Successful completion of 
these tests will allow Indian Head 
to recommend approval for release 
to service. The tests that will be per 
formed as part of the qualification 
effort are: 



- 147 - 



Visual inspection 

Radiographic Inspection 

Bridgewire integrity 

Electrostatic discharge 

Stray voltage 

Forty foot drop 

Six foot drop 

Temperature, humidity, and altitude 

cycling 

Salt fog 

Cook-off 

High temperature exposure 

High temperature storage 

Low temperature (-65°F) testing 

Ambient temperature (+70°F) testing 

High temperature (+200°F) testing 

APPLICATIONS 

The potential applications of RFAC 
include essentially all EEDs which 
must operate in HERO and ESD 
environments. We expect the re- 
duced cost, made possible by inte- 
grated circuit technology, will sub- 
stantially expand these applications 
in both the U.S. and U.K. However, 
In closing, I would like to discuss 
one novel application that may be 
of interest to this audience. 

The increasing availability of high 
power diode lasers has sparked the 
interest of the ordnance commu- 
nity. A fiber optic cable can conduct 
the optical energy into an initiator 
which is totally immune to RF and 
ESD hazards. Offering very 
lightweight and potentially low 
cost for safing and arming func- 
tions, the diode laser is nearly ideal 
for many applications. There is one 
catch; the diodes laser generates the 
optical energy with a low voltage 
(typically 3 volts) source. This cre- 
ates a single point failure safety 
problem unless mechanical means 
are used to block the light. How- 
ever, using an RFAC to isolate the 



diode power supply makes this 
problem disappear. A system which 
protects the diode and its power 
supply inside a Faraday closure 
should meet all safety require- 
ments for diverse applications such 
as crew escape systems, rocket mo- 
tor arm fire devices and automo- 
tive air bag initiation. 

BIOGRAPHIES 

Mr. Blachowski has held his pre- 
sent position as an Aerospace En- 
gineer in the Cartridge Actuated 
Devices Research and Develop- 
ment Branch at the Indian Head 
Division, Naval Surface Warfare 
Center for 5 years. In that time, he 
has been involved in exploratory 
R&D, advanced development, 
product improvement, and pro- 
gram support for cartridges and car- 
tridge actuated devices throughout 
the Navy. Mr. Blachowski received 
his Bachelor of Science degree in 
Aeronautical and Astronautical 
Engineering from Ohio State Uni- 
versity in 1985. 

As division vice president at Quart- 
tic Industries, Inc., Mr. Willis is re- 
sponsible for directing internal re- 
search and development activities, 
developing new products and new 
business activities. Product areas 
include electronic, electromechani- 
cal and ordnance devices used for 
safety and control of systems using 
energetic materials and processes. 

Mr. Willis received his Master of 
Science Degree in Physics from Yale 
University in 1959 and a Bachelor 
of Arts Degree from Wabash Col- 
lege. 



- 148 - 



514-28 
£1 14- 



Cable Discharge System for Fundamental Detonator Studies 31 



Gregg R. Peevy and Steven G. Barnhart 
Explosives Components Department 

William P. Brigham 
Explosives Projects and Diagnostics Department 

Sandia National Laboratories 
Albuquerque, NM 87185 



/o> 



ABSTRACT 

Sandia National Laboratories has recently completed 
the modification and installation of a cable discharge 
system (CDS) which will be used to study the 
physics of exploding bridgewire (EBW) detonators 
and exploding foil initiators (EFI or slapper). Of 
primary interest are the burst characteristics of these 
devices when subjected to the constant current pulse 
delivered by this system. The burst process involves 
the heating of the bridge material to a conductive 
plasma and is essential in describing the electrical 
properties of the bridgewire foil for use in 
diagnostics or computer models. The CDS described 
herein is capable of delivering up to an 8000 A pulse 
of 3 |is duration. Experiments conducted with the 
CDS to characterize the EBW and EFI burst 
behavior are also described. In addition, the CDS 
simultaneous VISAR capability permits updating the 
EFI electrical Gurney analysis parameters used in 
our computer simulation codes. Examples of CDS 
generated data for a typical EFI and EBW detonator 
are provided. 



♦This work was supported by the United States Department of 
Energy under Contract DE-ACO4-94AL85000. 



1.0 INTRODUCTION 

This paper describes the Cable Discharge System 
(CDS) and its use in fundamental detonator studies. 
The CDS is preferred over a conventional capacitor 
discharge unit (CDU) that delivers a decaying 
sinusoidal current pulse. The fast rising constant, 
"stiff', current, provided by the CDS charged 
cable(s) eliminated the uncertainty of a continuously 
changing current density that comes from a CDU. 
The CDS is actually two systems; the cable discharge 
system which provides a square wave current pulse 
to the detonator and the instrumentation system 
which measures the detonator parameters of interest. 
Fundamental detonator studies using the CDS 
generates information to be used in diagnostics or 
computer models. Computer modeling provides 
electrical/mechanical performance predictions and 
failure analysis of exploding foil initiator (EFI) and 
exploding bridgewire (EBW) detonators. This 
project is being performed in order to improve 
computer modeling predictive capabilities of EFI and 
EBW detonators. Previous computer simulations 
predicted a much higher voltage across the bridge 
than was measured experimentally. The data used in 
these simulations, for the most part, was collected 
two decades ago. Since this data does not adequately 
predict performance/failure, and instrumentation and 
measurement methods have improved over the years, 
the gathering of new data is warranted. 



- 149 - 



2.0 SYSTEM DESCRIPTION 

The cable discharge system (CDS) resides at Sandia 
National Laboratories New Mexico in Technical 
Area II and consists of the following hardware: 

■ Four 1000 foot long rolls of RG218 coaxial cable 

■ A high-voltage power supply (100 kV, 5 mA) 

■ A gas pressurized, self-breaking switch 

■ A gas system for pressurizing and venting the 
switch 

■ Custom output couplings with integral current 
viewing resistor (CVR) 

• Flat cable coupling for testing of exploding 
foil initiators (EFI) 

• Coaxial coupling for testing of exploding 
bridgewires (EBW) 

■ Instrumentation for measuring: 

• System current - current viewing resistor 
(CVR) 

• Voltage across the EBW/EFI bridge elements 
- voltage probes 

• Free-surface velocity of flying plate and 
particle velocities at interfaces for 
determining device output pressure - velocity 
interferometer system for any reflector 
(VISAR) 1 

■ Tektronix DSA602A digitizers 

■ 486DX33 PC 

The CDS is operated by pressurizing the output 
switch with nitrogen, charging the cables up to a 
pre-determined voltage which will deliver the 
required current to the device being tested when the 
switch is operated. The switch is operated by 
venting the gas with a fast-acting solenoid valve. 
Current from the CVR is used as a trigger source for 
the data recording system. A photograph of the CDS 
is given in Fig. 1 and a schematic of the CDS is 
given in Fig. 2. 

3.0 CAPABILITIES 

The four 1000 foot long RG218 coaxial cables can be 
configured to provide a current pulse ranging in 
amplitude from 100 to 8000 A with a width of 3 \xs 
and a risetime of 25 - 35 ns, Fig. 3 and 4. This wide 
range of current is made possible by the parallel 
connection of one to four cables. 

The system current is measured with a series 0.005 
Q CVR that is integral to the output of the CDS. A 
voltage probe is used to measure the voltage drop 
across the exploding element, either a bridgewire in 
the case of the EBW or the foil of an EFI (slapper). 



The voltage probe is a 1000 CI resistor placed in 
parallel with the bridge and a Tektronix CT-1 
current viewing transformer which measures through 
it. This allows for decoupling the measurements 
from ground and minimizes the possibility of ground 
loop problems. 

The VISAR is used to measure the free-surface 
velocity of the flyers of an EBW or EFI. It also can 
be used to measure particle velocity at a window 
interface which in turn, through the use of Hugoniot 
curves, can determine the explosive output pressure 
of an EBW. Two separate and independent VISAR 
modules can make these measurements 
simultaneously. They have different sensitivities 
therefore giving a high degree of confidence in these 
measurements. 

All three of these measurements (current, voltage, 
and velocity) are recorded on Tektronix DSA602A 
digitizing signal analyzers. These instruments have 
high bandwidth (up to 1 gHz) and high sampling 
rates (1 gHz for 2 channels of data). They also can 
produce calculated waveforms from basic current 
and voltage measurements representing: 

■ Resistivity 

■ Specific action 

■ Energy 

These calculated waveforms are produced from the 
measured data automatically as soon as they are 
recorded on the digitizer, see Fig. 5. 

The 486DX33 PC can acquire up to 16 waveforms 
on any shot and store it on a 90 megabyte Bernoulli 
disk. Other custom software does VISAR data 
reduction/analysis to give profiles of: 

■ Flyer velocity vs. time 

■ Flyer displacement vs. time 

■ Flyer velocity vs. flyer displacement 

Other commercial or custom software packages can 
then be used to create tables and/or graphs for 
presentation and report format. 

4.0 FUNDAMENTAL DETONATOR STUDIES 

This section briefly states the basic electrical/ 
mechanical theory of EBW and EFI detonators. For 
a more detailed explanation see the referenced 
reports 2 through 6. The CDS can be used to 
observe both electrical and mechanical behavior of 
detonators. From these observations, models with 
their parameters can be generated. Tucker and 
Toth 2 demonstrated that exploding wire (and foil) 



- 150 - 



resistivity, p, at fixed current density, j, may be 
uniquely specified as a function of either of two 
parameters: energy, e, or specific action, g. These 
relationships and equations are summarized below. 



p = /(e) or /(g) 



(1) 



The resistivity of the bridge is the voltage gradient 
across the bridge divided by the current density 
through the bridge. The resistivity is characteristic 
of the bridge metallic material. Resistance of the 
metallic bridge can be determined by accounting for 
the bridge volume; cross sectional area, A, and 
length, C 



R = pr/A 



(2) 



The energy deposited to the bridge is the integral of 
the voltage, V, and the current, i, over time. 



e = |Vidt 



(3) 



The specific action deposited to the bridge is the 
integral of the current density squared over time. 



g-jj 2 dt or (l/A 2 )Ji 2 dt 



(4) 



The characteristic resistivity versus specific action 
curve shows the resistivity of the bridgewire (foil) as 
it passes through the material phase changes; solid, 
liquid, and vapor, Fig. 6. Bridge burst is the 
condition in which the bridge is vaporized and arc 
breakdown occurs through the vapor. This 
corresponds to the peak resistivity. 

The characteristic resistivity as a function of energy 
or specific action curve is obtained from the 
instrumentation system by the measurement of the 
current through the bridge and the voltage drop 
across the bridge, Fig. 7. The voltage drop across 
the bridge divided by the current through the bridge 
gives the bridge resistance. Resistivity is calculated 
by multiplying the resistance by the bridge cross 
sectional area and dividing by the bridge length. 
Specific action is calculated by squaring the current 
and integrating over time while dividing by the 
square of the bridge cross sectional area. Energy is 
calculated by multiplying voltage and current and 
integrating over time. 

Tucker and Stanton 3 extended the Gurney method of 
predicting the terminal velocity of explosively driven 
projectiles to flyers driven by exploding foils in an 



EFI detonator. This couples input electrical 
parameters with EFI detonator mechanical output 
parameters; namely flyer velocity. The Gurney 
analysis of an explosive system is based on 
conservation of momentum and the assumption that 
the kinetic energy of the system is proportional to the 
total energy released by the exploding foil. The 
following is an approximate or simplified analysis 
(known as a modified Gurney analysis). The ratio of 
the system kinetic energy to the energy released is 
the Gurney efficiency, r\, and the Gurney energy, Eg 
is given by 



Eg^e 



(5) 



where e is the energy deposited into the foil. The 
solution of the momentum and energy equation 
yields a prediction of the terminal flyer velocity, Uf 



u f = (2 Eg) - 5 /(geometry) 



(6) 



where /(geometry) is a known factor dependent upon 
the EFI geometry. 

Tucker and Stanton 3 also showed that the Gurney 
energy could be empirically related to the burst 
current density of the foil. 



E g = kj b n 



(7) 



where k is the Gurney coefficient and n is the 
Gurney exponent. Measurement of the burst current 
density and knowledge of the Gurney energy allows 
the calculation of the Gurney coefficient and 
exponent. Once the Gurney coefficient and exponent 
are known, the terminal flyer velocity can be 
calculated for a determined burst current density. 

From the measurement of the flyer velocity, the flyer 
pressure pulse magnitude, P, and duration, t, 
imparted to the explosive receptor can be calculated 
from Hugoniot pressure - particle velocity (P-u) 
relationships. 4 Explosive initiation criteria can then 
be calculated to determine initiation margin. Some 
explosives are characterized by the relationship 



P n x>K 



(8) 



where K is a constant and n is an exponent specific 
to an explosive. 5 For explosive initiation 
(detonation), this criterion must equal or exceed the 
specified constant, K. Other explosives are 
characterized by initial shock pressure, P, versus run 



- 151 - 



distance to detonation, x, plots or "Pop plots". 6 
These plots can be expressed empirically by the 
relationship 



log P = A - B log X or 
P = C + Dx" 1 



(9) 
(10) 



where A, B, C, and D are constants for least squares 
data fit. 

4.1 SAMPLE DATA 

A typical EBW and EFI detonator were selected and 
tested in order to present typical data output. The 
EBW is a 1.2 x 20 mil Au bridgewire in contact with 
a PETN explosive DDT column (18.5 mg 0.88 
g/cm 3 initial pressing, 9.3 mg 1.62 g/cm 3 output 
pellet). The EFI is a 15 x 15 x 0.165 mil square Cu 
bridge with a 1 mil thick Kapton flyer. An 
investigation was conducted to observe the behavior 
of the detonators over a range of operating current 
density. The response of the EBW and EFI 
detonators to a similar CDU burst current level pulse 
was also investigated to verify that CDU and CDS 
generated data are comparable. The data is 
presented in a summary format at bridge burst 
condition. Bridge burst resistivity, specific action, 
and energy are plotted versus current density, Fig. 8 
through 13. 

Based upon these results, several observations could 
be made. 

■ Resistivity at bridge burst decreases with 
increasing current density for the EBW, Fig. 8. 

■ Specific action to bridge burst remains fairly 
constant over current density range for both the 
EBW and EFI, Fig. 9 and 12. 

■ Energy to bridge burst increases with increasing 
current density for both the EBW and EFI, Fig. 
10 and 13. 

■ Resistivity at bridge burst remains fairly 
constant over current density range for the EFI, 
Fig. 11. 

■ CDU and CDS generated data are comparable, 
Fig. 11 through 13. 

From these observations, in order to adequately 
model an EBW or EFI detonator, resistivity versus 
specific action or energy data needs to be taken at 
three current levels; slightly above threshold (50% 
fire/no-fire), 1.5 times threshold, and 2 times 
threshold (normal minimum operation). 

Previous computer simulations always predicted a 
much higher voltage across the bridge than was 



measured. The "old" resistivity versus specific 
action look-up table data for a gold EBW and a 
copper EFI are compared to recently, "new", 
generated data in Fig. 14 and 15. Since the peak 
resistivity is much lower, predicted voltages now 
closely match actual values. 

The mechanical output of both the EBW and EFI 
detonators was also observed over the range of 
operating current density by using the VISAR. The 
flyer velocity of the EBW closure disk was measured 
at a PMMA window interface. Flyer velocity did not 
change over the current range. This was expected, 
Fig. 16. EFI flyer velocity increased with increasing 
current density as expected, Fig. 17. The EFI 
Gurney energy was calculated from the flyer velocity 
and current density, Fig. 18. A least squares data fit 
yielded a Gurney coefficient of 0.00125 and a 
Gurney exponent of 1.28. 

5.0 INTEGRATION OF DETONATOR 

STUDIES DATA INTO PREDICTIVE 
COMPUTER CODES 

Detonator studies data are used in computer codes to 
predict the electrical and mechanical behavior of 
bridges as they burst. An electrical circuit solver 
computes current as a function of time and then 
"looks up" resistivity in a look-up table or calculates 
the resistivity with an empirical relationship as a 
function of computed energy or specific action to get 
the instantaneous bridge resistance. Once a burst 
current is known, for an EFI, a flyer velocity can be 
calculated. A list of some of the electrical circuit 
solvers available along with a brief description 
follows. 

■ AITRAC - complex circuit solver with bursting 
wire & foil look-up table^ 

■ CAPRES - simple circuit solver with bursting 
foil look-up table and electrical Gurney routine 

■ PSpice© - complex circuit solver with bursting 
elements added by Furnberg (empirical 
function) and Peevy (look-up table) with 
electrical Gurney and P 11 * initiation criterion^ 

■ Fireset - code by Lee with empirical resistivity 
function 10 

■ Slapper - code by R. J. Yactor, Los Alamos 
National Laboratories, with empirical resistivity 
function, electrical Gurney, and Pop plot 
explosive initiation criteria. 

Electrical Gurney and P^ initiation criterion are 
implemented into the PSpice electrical circuit 



- 152 - 



simulator as custom circuit elements using the 
Analog Behavioral Modeling capability. 

5.1 COMPUTER PREDICTION VERSUS DATA 

A computer simulation of a typical capacitor 
discharge unit (CDU) firing system with a single EFI 
detonator was performed using the latest CDS 
generated resistivity versus specific action data look- 
up table. The firing system lumped parameters are: 

C = 0.2^F 

R = 100 mQ 

L = 17 nH. 
Simulation output versus test data is shown 
graphically in Fig. 19 and 20. As can be seen in 
Table 1, simulations compare well to experiment. 

6.0 CONCLUSION 

The CDS has been fully documented. Newly 
obtained data are more accurate and improves 
computer simulation, electrical/mechanical 
performance predictions and failure analysis of EBW 
and EFI detonators. Future plans are to model other 
EBW and EFI detonators of interest. 



6. B. M. Dobratz, P. C. Crawford, LLNL 
Explosives Handbook Properties of Chemical 
Explosives and Explosive Simulants. Lawrence 
Livermore National Laboratory, Report No. 
UCRL-52997, 1985. 

7. Berne Electronics Inc., Sandia National 
Laboratories. Albuquerque. AITRAC 
Augmented Interactive Transient Radiation 
Analysis bv Computer User's Information 
Manual . Sandia National Laboratories, Report 
No. SAND77-0939. 

8. PSpice (a registered trademark of) Microsim 
Corporation, 20 Fairbanks, Irvine, California. 

9. G. R. Peevy, S. G. Barnhart, C. M. Furnberg, 
Sla pper Detonator Modeling Using the PSpice© 
Electrical Circuit Simulator. Sandia National 
Laboratories, Report No. SAND92-1944. 

10. R. S. Lee, FIRESET. Lawrence Livermore 
National Laboratory, Report No. UCID-21322, 
1988. 



7.0 REFERENCES 

1. O. B. Crump, Jr., P. L. Stanton, W. C. Sweatt 
The Fixed Cavity VISAR. Sandia National 
Laboratories, Report No. SAND-92-0162. 

2. T. J. Tucker, R. P. Toth, EBW1: A Computer 
Code for the Prediction of the Behavior of 
Electrical Circuits Containing Exploding Wire 
Elements . Sandia National Laboratories, Report 
No. SAND-75-0041. 

3. T. J. Tucker, P. L. Stanton, Electrical Gurnev 
Energy: A New Concept in Modeling of Energy 
Transfer from Electrically Exploded 
Conductors . Sandia National Laboratories, 
Report No. SAND75-0244. 

4. P. W. Cooper, "Explosives Technology Module 
D, Shock and Detonation," Sandia National 
Laboratories Continuing Education in Science 
and Engineering, INTEC Course No. ME717D. 

5. A. C. Schwartz, Study of Factors Which 
Influence the Shock-Initiation Sensitivity of 
Hexanitrostilbene (HNS) . Sandia National 
Laboratories, Report No. SAND80-2372. 



- 153 - 




Fig, 1 Photograph of Cable Discharge System 



— (heo) D=-U}- 



( VBOTT \ / 1B88FT A 
I RG218 J I R0218 J 





J>: 



GAS 
WITO 



i»*v 
powet 

SUPPLT 
















TOO 


PULSE 
GENERATOR 




DELAY 








GENERATOR 




Fig. 2 Schematic of Cable Discharge System 



- 154 - 



DSA 602 A DIGITIZING SIGNAL ANALYSER 
date: 5 -JAN- 9 4 tia«: 10:52:01 

Tfek 



1.125V 



D 



125mV 




CABLE DISCHARGE SYSTEM 
•OUTPUT INTOSHORT CIRCUIT- 




Fig. 3 Typical DSA Current Waveform 



DSA 602A DIGITIZING SIGNAL ANALYZER 
data: 5-JAN-94 tim»z 10:55:13 

Ttek 



-12SiiV- 



-ll.Bn« 
7.68 









! L-'M^T'I HI 




125m\ 


: J •' '• ''■ : '■ = -' 
: \.l CABLE DISCHARGE SYSTEM 

/ OUTPUT INTO SHORT riPriITT \ 


LI 






_J/ ^ ! ; \ \ \ \ \ I 


X' 





-1BBB AMPSi 



lBns^dlv ET} 



B8.2n» 



.anHEHrcHM 



Fig. 4 Typical Current Leading Edge 



- 155 - 



DSA 602A DIGITIZIHG SIGNAL ANALYZER 
dot*: 6-AUG-93 tla*: 13:20:30 



Ttek 



675mUj 




Fig. 5 Calculated Waveforms 



6.00E-04 T 




O.OOE+00 



O.OOE+00 5.00E+08 1.00E+09 1.50E+09 

Specific Action (A A 2-s/cm A 4) 



2.00E+09 



1,2x20® 400 A 



Fig. 6 Typical Resistivity vs. Specific Action Profile 



- 156 - 




200 300 

TIME(nSEC) 



-CURRENT 



Ufend 
VELOCITY 



- VOLTAGE 



Fig. 7 Cable Discharge System Waveform for EFI 



1000 



900 



800 



E 700 



o 




§ 


600 


g 


500 


£ 


400 


£ 




CO 


300 


CO 




UJ 




cc 


200 




100 



■r ~ "* " " " " ~ — " ~~ " 

^ 

■ 
1 «H , m 

_j_j l, , 



20 40 60 80 100 120 140 160 
CURRENT DENSITY (amps/cm ~ 2) 
(Millions) 



180 200 



Fig. 8 EBW Burst Data Summary of Resistivity vs. Current Density 



- 157 - 



< 



CM ~ 
< °* 



9 $ 

o 



o 

LU 

a. 

CO 



1.8 
1.6 
1.4 
1.2 
1 
0.8 
0.6 
0.4 
0.2 



— *— * - * -*-j - u- 

■ 

_ __ «. __ I _,..„ _ _ 

— _ j_ 1 



20 40 60 80 100 120 140 160 180 200 
CURRENT DENSITY (amps/cm ~ 2) 
(Millions) 



Fig. 9 EBW Burst Data Summary of Specific Action vs. Current Density 



D 
O 



> 

o 
cc 

W 

2 
LU 



100 
90 
80 
70 
60 
50 
40 
30 
20 
10 



' ' T m ir " l - 

_ „*„■ „ ||f J L _ 

- - ■■— - — — 



20 40 60 80 100 120 140 160 

CURRENT DENSITY (amps/cm ^ 2) 
(Millions) 



180 200 



Fig. 10 EBW Burst Data Summary of Energy vs. Current Density 



- 158 - 



2.50E-04 



2.00E-04 



E 
V 

J 1.50E-04 

O 



> 

P 1.00E-04 

55 

UJ 



5.00E-05 



O.OOE+00 



• # 



• CDS DATA 
■ CDU DATA 



O.OOE+00 5.00E+06 1.00E+07 1.50E+07 2.00E+07 2.50E+07 J.00E+O7 

CURRENT DENSITY (amps/cm A 2) 



3.00E+09 



T 2.50E+09 

E 
o 



CM 

< 



2.00E+09 



•2* 1.50E+09 

Z 

o 

H 

O 

< 1.00E+09 

O 



O 
UJ 

£ 5.00E+08 

(0 



O.OOE+00 



Fig. 11 EFI Burst Data Summary of Resistivity vs. Current Density 






> CDS DATA 
i CDU DATA 



O.OOE+00 5.00E+06 1.00E+07 1.50E+07 2.00E+07 2.50E+07 3.00E+07 

CURRENT DENSITY (amps/cm A 2) 



Fig. 12 EFI Burst Data Summary of Specific Action vs. Current Density 



- 159 - 



150.0 



120.0 - 



3 90.0 - 

o 



g 60.0 -f 



• CDS DATA 
° CDU DATA 



30.0 -- 



0.0 



O.OOE+00 



1 

5.00E+06 



-h 



-h 



H 



1 .00E+07 1 .50E+07 2.00E+07 2.50E+07 3.00E+07 

CURRENT DENSITY (amps/cm A 2 



Fig. 13 EFI Burst Data Summary of Energy vs. Current Density 



1.20E43 T 



^ 1.00E-03 - 

E 
V 

g 8.00E-04 - 

^ 6.00E-04 -- 

> 

to 4.00E-04 



2.00E-04 - 



0.00E+00 \ \ ><** * 




1.2x20 ©400 A 
Tucktrdata 



i .i p~ — , ■ ~ — . -J — t- m ft » m — 

0.00E+00 1.00E+09 2.00E+09 3.00E+09 

Specific Action (A A 2-s/cm A 4) 



4.00E+09 



Fig. 14 New vs. Old EBW "Look-up" Table 



- 160 - 




O.OOE+00 



O.OOE+00 4.00E+09 8.00E+O9 1.20E+10 1.60E+10 

Specific Action (A A 2-s/cm A 4) 



15x16x.165e3kA 

Old look-up tabte 



2.00E+10 



Fig. 15 New vs. Old En "Look-up" Table 



■t i r 



> i i i 



Orertay 



i i i i 




Nv 



'^^^*^^ 



~ +T~° 



°4— - 



1! %% 



200 300 

TIME(iiSECONDS) 



- VELOCITY, 6KA 
VELOCITY, 4KA 



Legend 

■ VELOCriY, 8KA 



- VELOCITY, IKA 



Fig. 16 EBW Output Flyer Velocity vs. Current Density 



- 161 - 



« -r 



4 -- 



E 3 

E 



8 2 



1 " 



-h 



H- 



-+- 



PART OF CURRENT SHUNTED 
AWAY FROM BRIDGE 



• COS DATA 

* CDUDATA 



600 1000 1500 



2000 2500 3000 3500 

CURRENT (amps) 



4000 4500 5000 



Fig. 17 EFI Flyer Velocity vs. Current Density 






>% 



c 

UJ 

ST 

c 

3 

o 



ivv - 
















































































































































y 

' m 
















10 - 


/ 

m 
















1 




















1 



■ .5*v A 2/fo*o 
k*j*nFlt 



1000 



10000 



Burst Current Density (GA/m A 2) 



Fig. 18 EFI Gurney Energy vs. Burst Current Density 



- 162 - 



4.0K T 



l.OK 




Os lOOna 

- I(RFS) • V(12) 



Ti» 



Fig. 19 Simulation Using New Look-Up Table Voltage and Current Traces 



lMt.(! : Hl/Hh/!M 





























A 






















/ 


y 




















/ 


\ 




















! 


\ 






















\ 


\ 






















\ 






















\ 


s 






















\ 












II 


1 M 


i n 


1 11 


i ii 


A II 


i li 


i <n 


i fl 


■ ii 


i 



TM<n$> 



Fig. 20 Test Data Voltage and Current Traces 



- 163 - 



Table 1 
EFI Simulation Output vs. Test Data 



CDU 
Charge 
Voltage 

(V) 


Calculated 

Burst 

Current 

(A) 


Calculated 

Flyer 

Velocity 

(mm/us) 


Burst 
Current 

(A) 


Flyer 
Velocity 

(mm/us) 


1950 


3009 


4.1 


3020 


4.2 


2600 


4072 


4.8 


3960 


4.9 


2800 


4325 


5.0 


4170 


5.1 



- 164 - 



INITIATION CURRENT MEASUREMENTS FOR HOT BRIDGEWIRE DEVICES 

Gerald L. O'Barr (Retired, General Dynamics, 1 Oct 1993) 






) ^ 



ABSTRACT One-shot type testing of hot bridgewire 
explosive cartridges provides the weakest possible firing 
characteristic data. One-shot type testing includes the 
Bruceton and the Probit methods, and all their off-shoots. 
One-shot testing is used for only one reason: the fear of 
"dudding." Modern programmable power supplies and 
oscilloscopes can now be used to obtain data with no fear of 
dudding. Any continued use of these old, one-shot type test 
methods is irresponsible behavior. Our society deserves 
better testing methods and will get them through the courts 
if we cannot make these changes on our own. 



1 . INTRODUCTION 

Explosive cartridges often use 
a bridgewire initiation 
system. A bridgewire is a 
very small wire, approximately 
0.005" in diameter and 0.2" 
long, through which electrical 
current can be forced to flow. 
Because of the electrical 
resistance of the wire, the 
current can cause the wire to 
become hot like the filament 
in a light bulb. Around the 
bridgewire is a sensitive 
ignition mix that will ignite 
with temperature. As the 
bridgewire gets hot, it 
ignites the ignition mix, 
which then sets off the main 
charge of the cartridge, often 
through intermediate mixes 
between the ignition mix and 
the main charge. 

The reliability of operation 
of an explosive device is 
critical. Being a destructive 
event, one would not want an 
explosive cartridge to go off 



before it is needed (for 
safety) and yet, when its 
function is required, since 
explosive devices are usually 
used for critical operations, 
it must work quickly and 
reliably. Since the primary 
initiation is by the flow of 
current, the following 
questions must be asked: 

A. What is the maximum 
current that can flow 
through the bridgewire 
and not have it ignite or 
explode? 

B. What is the minimum 
current that can be used 
and still be sure that it 
will ignite or explode? 

The answer to A provides what 
is called the "no-fire" 
current. It tells one the 
degree of safety that might 
exist in using this device. 
If the current for A is so low 
that random stray voltages 
might induce sufficient 



- 165 - 



current to set it off, then it 
would be unacceptable. Often, 
a one-ampere current is 
specified as being the minimum 
acceptable no-fire current. 
If the bridgewire resistance 
could be less than one ohm, a 
power minimum of one watt is 
also sometimes specified. 

The answer to B provides the 
"all-fire" current. This 
value must obviously be more 
than A, and hopefully low 
enough that the power supply 
being used to set it off can 
provide the current that is 
required. Values of 3.5 to 5 
amperes are often specified. 
In statistical terminology, 
the all-fire and no-fire 
current requirements are often 
stated as follows: 

A. The cartridges shall have 
a 1 ampere/1 watt or 
greater no-fire current 
with a reliability of 
99.9%, at a confidence 
level of 95%. 

B. The cartridges shall have 
a 3.5 ampere or less all- 
fire current with a 
reliability of 99.9%, at 
a confidence level of 
95%. 

These statistical 
requirements can be determined 
if the firing currents are 
normally distributed and the 
mean and the standard 
deviation of the lot's firing 
current can be determined. 
Basically, this report will 
discuss a new test method for 
measuring the mean and the 
standard deviation of the 
firing current for a lot of 



explosive cartridges. 

2.0 DIFFICULTIES IH MEASURING 
THE INITIATION CURRENT 

The phrase "firing current" 
has many different meanings. 
This results in certain 
difficulties that we will take 
care of now. The firing 
current is often used to mean 
the current being applied to 
the bridge-wire. It is also 
used for the minimum current 
required to ignite a 
cartridge. It is this minimum 
current, along with its 
distribution, that allows us 
to determine the all-fire and 
the no-fire. Therefore, when 
we mean this minimum current, 
we shall use the words, 
"initiation current." 

To measure the initiation 
current, it would seem easy to 
connect a cartridge up to a 
variable power source and 
manually turn up the current 
until it ignites. Another 
method would be to use a 
series of ever increasing 
steps or pulses of current. 
The current at which it 
ignites could then be the 
desired data. 

When cartridges were 
originally made (over 30 years 
ago), such efforts often 
proved to be impossible. The 
"sensitive" ignition mix, when 
exposed to heat, could 
degrade. If one raised the 
current too slowly, or exposed 
a cartridge to several firing 
currents below that required 
for ignition, the cartridge 
could actually become 
impossible to ignite. When 
the chemicals making up the 



- 166 - 



ignition mix become 
sufficiently degraded, the 
cartridge becomes a dud. 
Thus, the initiation current 
was often dependent on the 
rate at which the current was 
applied, and could become 
infinite. The industry 
quickly learned that only 
"virgin" cartridges could be 
tested in a one-time-only 
test. 

For this reason, all tests for 
explosive cartridges (Bruceton 
testing, Probit testing, and a 
multitude of testing based 
upon these approaches) use a 
one-shot approach. A 
cartridge is exposed to one 
pre-selected current value, 
and a fire or no-fire result 
is recorded. 

The data obtained by this 
approach is not good. If a 
cartridge is tested at 2.30 
amperes and it fires, no one 
can say that this was the 
initiation current for this 
cartridge. The Bruceton and 
Probit test methods would use 
this data as a firing point, 
but all one knows for sure is 
that the true initiation 
current for this cartridge was 
this value or less. It could 
have been much less. It could 
have been so much less that it 
could have skewed all the rest 
of the data. But in the one- 
shot method, one will never 
know the true initiation 
current for that particular 
cartridge. 

The same things are true if 
the cartridge did not fire. 
Again, the Bruceton and the 
Probit test methods assume 



that it is the no-fire point. 
The actual no-fire point might 
be much greater. It could 
even be a dud, with an 
infinite firing current value. 
Neither the Bruceton nor the 
Probit methods will be 
sensitive to these situations. 

Because the data that comes 
from a one-shot test is not 
normally the initiation 
current for that cartridge (if 
it fires), nor the real no- 
fire current for that 
cartridge (if it does not 
fire), then the data is 
extremely poor data. It has 
almost no meaning except as 
limits to the values being 
sought, and these limits have 
no value unless they are 
repeated a great number of 
times in a well controlled 
lot. If you are making tests 
on a lot that is not well 
controlled, the limit data 
will be entirely useless. 
Worst of all, one can seldom 
tell by looking at the 
Bruceton or Probit data if the 
data is from an adequately 
controlled lot, or when the 
number of limit tests are 
really adequate. 

3.0 NEW METHODS 

Today, one jjoes not need to 
manually turn up the current 
to fire a cartridge. One does 
not need to follow a moving 
indicator to read what might 
have been the initiation 
current . Today , with 
programmable current 
generators, and oscilloscopes, 
very controlled current 
profiles and measurement 
devices can measure the actual 
initiation current of a 



- 167 - 



device. It can be done at a 
rate where no significant 
degrading will occur. 

Thus, today, in any modern 
laboratory, there is no need 
to work with the out-of-date, 
and very poor data generating 
methods such as the Bruceton 
or Probit one-shot methods. A 
"dynamic ramp" test method can 
supply reliable and specific 
initiation data for all 
cartridges tested. 

It is also a fact that over 
these many years, much more 
stable ignition mixes are now 
being used. For most modern 
explosive cartridges, one 
could probably even use a 
manual method and obtain 
better data than what a 
Bruceton or Probit method 
might provide. 

4.0 THE DYNAMIC RAMP TEST 
METHOD The dynamic ramp 
test used a programmable power 
supply that puts out a 
controlled current ramp from 
zero to five amperes. The 
rate of the ramp, as used in 
these tests, was approximately 
3.5 amperes per second. The 
current circuit included a 
standard one-ohm resistor in 
series with the explosive 
cartridge. The voltage across 
the one-ohm resistor was 
recorded on an oscilloscope 
with memory trace recording. 
The scope was calibrated so 
that within the expected 
firing levels, currents could 
be read to the nearest 0.01 
amperes, with an overall 
accuracy estimated to be 
better than ±0.02 amperes. 



The firing point of the 
cartridges being tested 
appeared to be a clean break 
in the circuit, with the 
voltage returning to zero when 
the bridgewire initiated the 
cartridge. If the 
programmable power supply has 
too high of a voltage 
capability, current can 
continue to flow even when the 
bridgewire is "broken" by 
flowing the current through 
the ionized gas that is 
created in an explosion. This 
would indicate a higher firing 
current then that actually 
required. When ionization 
flow is present, the trace is 
usually very uneven without a 
clean break. 

5.0 RESULTS 

The results for a series of 
ten consecutive firings are 
shown in Table I. The results 
from a small Bruceton test of 
17 firings is shown in 
Appendix A. 

The means determined from 
these two tests were almost 
identical, 2.43 amperes in the 
dynamic ramp test and 2.44 
amperes in the Bruceton. The 
standard deviations, however, 
are much different, 0.234 
amperes for the dynamic ramp 
test and only 0.070 amperes 
for the Bruceton test. This 
difference in the standard 
deviation is over three to 
one. 

Table 2 shows the results 
after 20 firings. The data 
continues to confirm the 
statistics that had been 
obtained in Table 1. Figure 
1 shows the distribution for 



- 168 - 



TABLE 1. DYNAMIC RAMP TEST 



PART NO. :55-06018-2 LOT 13-37700 DATE: 1 FEB 1993 





CARTRIDGE 
NO. 


PEAK 
CURRENT (Xi) 




(Xi-X) A 2 


1 


86845 


2.50 




0.0045 


2 


86811 


2.14 




0.0858 


3 


86881 


2.39 




0.0018 


4 


86850 


2.32 




0.0128 


5 


86854 


2.71 




0.0767 


6 


86864 


1.97 




0.2144 


7 


86991 


2.66 




0.0515 


8 


86883 


2.55 




0.0137 


9 


86866 


2.60 




0.0279 


10 


86816 


2.49 




0.0032 



SUM Xi = 


24.33 


SUM(Xi-X) A 2 = 


0.4924 


N = 


10 






t (for90%,n= 9) = 


1.83 






MEAN = X = 


SUMXi/ N 


— 


2.433 



STANDARD DEVIATION = S.D. = 

(SUM(Xi-X) A 2/(N-1)) A .5 = 0.234 

STANDARD ERROR 

OF THE MEAN (S.D.)/(N\5) = 0.074 

STANDARD ERROR 

OF THE S.D. = (S.D.)/(2N A .5) = 0.052 

ALL FIRE = 

X + t (S.D./N A .5) + 3.09 ( S.D. + t S.D./(2N)\5) = 3.587 

NO FIRE = 

X - t (S.D./N A .5) - 3.09 ( S.D. + t S.D./(2N)\5) = 1 .279 



- 169 - 



TABLE 2. DYNAMIC RAMP TEST 



PART NO. :55-06018-2 LOT 13-37700 DATE: 1 FEB 1993 





CARTRIDGE 
NO. 


PEAK 
CURRENT (Xi) 




(Xi-X) A 2 


1 


86845 


2.50 




0.0057 


2 


86811 


2.14 




0.0809 


3 


86881 


2.39 




0.0012 


4 


86850 


2.32 




0.0109 


5 


86854 


2.71 




0.0815 


6 


86864 


1.97 




0.2066 


7 


86991 


2.66 




0.0555 


8 


86883 


2.55 




0.0158 


9 


86866 


2.60 




0.0308 


10 


86816 


2.49 




0.0043 


11 


86826 


2.50 




0.0057 


12 


86907 


2.50 




0.0057 


13 


86815 


2.35 




0.0056 


14 


86863 


2.54 




0.0133 


15 


86888 


2.02 




0.1636 


16 


86885 


2.58 




0.0242 


17 


86814 


2.05 




0.1403 


18 


86829 


2.47 




0.0021 


19 


86956 


2.65 




0.0509 


20 


86926 


2.50 




0.0057 



SUM Xi = 


48.49 


SUM(Xi-X) A 2 = 


0.9101 


N = 


20 






t (for90%,n=19) = 


1.725 






MEAN = X = 


SUMXi/ N 


— 


2.425 



STANDARD DEVIATION = S.D. = 

(SUM(Xi-X) A 2/(N-1))\5 = 0.219 

STANDARD ERROR 

OF THE MEAN (S.D.)/(N A .5) = 0.049 

STANDARD ERROR 

OF THE S.D. = (S.D.)/(2N A .5) = 0.035 

ALL FIRE = 

X + t (S.D./N A .5) + 3.09 ( S.D. + 1 S.D./(2N) A .5) = 3.370 

NO FIRE = 

X - t (S.D./N\5) - 3.09 ( S.D. + t S.D./(2N)\5) = 1 .479 



- 170 - 



FIGURE 1. FIRING DISTRIBUTION 

55-06018-2 LOT 13-37700 1 FEB 1993 




D 



TOTAL DISTRIBUTION 
FIRST 10 
LAST 10 



1.85 1.95 2.05 2.15 2.25 2.35 2.45 2.55 

FIRING CURRENT (AMPERES) 



2.65 2.75 



2.85 



these firings. A deviation 
from a normal distribution 
does exist , but a larger set 
of data would be desirable 
before too much should be made 
of these details. The 
important point is made, the 
distribution function is 
obtainable by this test 
method . 

6.0 LIMITATIONS OF THE NEW 
METHOD There are only two 
limitations in using the 
dynamic ramp test method. The 
current ramp can be too slow 
or too fast. 

If one were to use a very fast 
current ramp (possibly over 30 
amperes per second) , a thermal 
diffusivity effect might be 
observed. This effect is due 
to the amount of time it takes 
the heat energy generated in a 
bridgewire to distribute 
itself out to the ignition mix 
to ignite it. If during this 
time, the current in the 
bridgewire changes, the 
initiation current will appear 
to be slightly greater than 
that actually required. 
Because bridgewires are so 
small, and are made of metals 
that have relatively high 
thermal conductivity values, 
this time is very small. 

This time can be estimated in 
other ways. The cartridges 
which we were testing normally 
fire in one millisecond when 
using twice their mean 
initiation current. 
Therefore, their time constant 
is about 3.5 milliseconds. 
Using a ramp rate of only 3.5 
amperes per second, an error 
of about .01 amperes would 



occur due to these diffusivity 
effects. 

The other limit is being too 
slow. The slower the ramp 
rate, the greater could be the 
degradation of the ignition 
mix. With a ramp rate of 3.5 
amperes per second, and all 
firings being at a value of 
less than 3.5 amperes, this 
means that all the units fired 
within one second of time. 
Very little degradation could 
be expected in conditioning 
times as short as one second. 

7.0 ADVANTAGES OF THE NEW 
METHOD Explosive cartridges 
are usually very expensive. 
In order to obtain reliability 
they must be made in large 
lots under controlled 
conditions, and a large 
percentage of the lot tested. 
The dynamic ramp test will 
directly reduce the number of 
cartridges required for 
testing. Even better than 
this, however, is a great 
increase in reliability. The 
dynamic ramp test will catch 
all cartridges that might be 
outside of the normal 
distribution. Bruceton and 
Probit type testing cannot 
produce data specific to a 
single cartridge, and 
therefore abnormal cartridges 
tested with these methods 
cannot make much, if any, 
change in the final results. 

The real power of the dynamic 
ramp test is providing, for 
the very first time, the 
specific initiation current 
for individual cartridges. 
This data will now make it 
possible to actually confirm 



- 172 - 



the true current distribution 
of a lot. 

8.0 CONCLUSION 

If one has modern test 
equipment, there are no known 
reasons why one would not want 
to use the dynamic ramp test 
method. It brings the testing 
of explosive cartridges back 
to standard statistics. 

Books have been written on the 
questionable assumptions and 
misuse of data from the 
Bruceton and Probit methods. 
All those questions disappear 
with the use of the dynamic 
ramp test. The applications 
of the Bruceton and Probit 
methods require special charts 
and calculations. The dynamic 
ramp test uses the same 
statistics that would be used 
in any other normal approach. 

It is true that the dynamic 
ramp test is still a 
statistical test. The 
meaningfulness of the results 
will, as always, depend upon 
how well the test units 
reflect the distribution of 
the lot. If non-random or 
incomplete selections are made 
from a lot, the dynamic ramp 
test data will be faulty to 
the extent that this occurs. 
But, this is true for all 
statistics, and can be guarded 
against in the same way that 
all other statistical tests 
are handled. 

This new method actually 
allows additional research. 
If at any time a set of 
cartridges are identified as 
being "out-of-family, " having 
seen some unexpected exposure 



or has some other observed 
anomaly, the dynamic ramp 
test, with very few test 
units, can quickly tell if 
those anomalies are causing 
any effect in their firing 
characteristics. For Bruceton 
or Probit testing, you might 
need 30 or 40 units before you 
could feel good as to whether 
a difference exists between 
two different groups. 

Research into thermal 
diffusivity effects and into 
degradation effects could also 
be easily accomplished. These 
reasons make the dynamic ramp 
test an exciting approach to a 
problem that has existed for 
over thirty years. It will be 
a powerful and necessary 
approach if we are to remain 
competitive. 

9 . RECOMMENDATIONS 

Government regulations 
requiring Bruceton or Probit 
testing of bridgewire type 
explosive cartridges should be 
modified to include, as a 
preference, the dynamic ramp 
testing approach. The dynamic 
ramp testing is a major 
improvement. It provides 
stronger, more direct 
statistical data for 
determining the mean and the 
deviation of the initiation 
current for bridgewire type 
explosive cartridges. In 
addition, the true 
distribution of the data is 
observable. For all previous 
testing methods, assumptions 
of the distributions were 
required and could never be 
actually confirmed. 
Therefore, all previous 
testing was always with some 



- 173 - 



uncertainty, especially where 
projections of several 
standard deviations were 
required. 

The dynamic ramp test method 
allows the initiation current 
of individual cartridges to be 
measured. This greatly 
increases the research that 
can be done with explosive 
cartridges. The effects of 
any manufacturing variable or 
any environmental exposures 
can readily be assessed. 
Also, by taking the ramp rate 
to great extremes, the 
diffusivity effects and the 
degrading effects (if any) of 
any particular design can be 
quickly determined. 



- 174 - 



APPENDIX A 

CALCULATIONS (55-06018-2, Lot 13-37700) 

Fires No-Fires (NO i x Ni i 2 x Nj 

10 

5 12 4 

3 4 4 4 

12 



Applied 


I i 


2.6 


3 


2.5 


2 


2.4 


1 


2.3 





Mean 


(XR) 


Xr 


= 



10 N = 7 A=6 B = 8 



10 + Al (A/N + 1/2) 
2.3 + 0.1 (6/7 + 1/2) 
= 2.44 amperes 

Standard Deviation (or) 
M = N x B - A 2 

N 2 
7 X 8 - 6 2 

7 2 
0.4082 

Therefore, 

S = 0.70 (from table) 

or = S(AI) 

(0.70) (0.1) 
0.070 

Sampling Error (o a ) 

a a = or H/N 1/2 (H from table) 

0.07 (1.7)/7 1/2 

.045 

Sampling Error (o x ) 

o x = or G/N 1/2 (G from table) 

0.07 (1.3)/7 1/2 

0.0344 

Confidence Intervals (single tail statistics) 

Y + ta v (t = 1.94 from table at 95% level of confidence) 

1. For the mean: Xr + to x 

2. For the standard deviation: or + to a 



- 175 - 



APPENDIX A - Continued 



No-Fire (0.999 reliability at 95% confidence) 

(Xr - tax) -3.09 (or + ta a ) > 1 ampere 

2.44 - 1.94 (0.0344) - 3.09 [0.07 + 1.94 (0.045)] > 1 ampere 

1.88 > 1 ampere 

All-Fire (0.999 reliability at 95% confidence) 

(Xr + tax) +3.09 (aR + ta a ) < 3.5 amperes 
2.44 + 1.94 (0.0344) + 3.09 [0.07 + 1.94 (0.045)] < 3.5 amperes 
2.99 < 3.5 ampere 



- I7B - 






?■ 



DEVELOPMENT AND DEMONSTRATION 

OF AN 

NSI-DERIVED GAS GENERATING CARTRIDGE (NGGC) 



by 



Laurence J. Bement 

NASA Langley Research Center 

Hampton , Virginia 

Morry L. Schimmel 

Schimmel Company 

St. Louis, Missouri 

Harold Karp 

Hi-Shear Technology 

Torrance, California 

Michael C. Magenot 

Universal Propulsion Co. 

Phoenix, Arizona 



Presented at the 1994 NASA Pyrotechnic Systems Workshop 

February 8 and 9, 1994 
Sandia National Laboratories 
Albuquerque, New Mexico 



- 1/7 - 



DEVELOPMENT AND DEMONSTRATION OF AN NSI-DERIVED 
GAS GENERATING CARTRIDGE (NGGC) 



Laurence J. Bement 

NASA Langley Research Center 

Hampton, Virginia 

Harold Karp 
Hi-Shear Technology 
Torrance, California 



Morry L. Schimmel 
Schimmel Company 
St. Louis, Missouri 

Michael C. Magenot 

Universal Propulsion Co. 

Phoenix, Arizona 



Abstract 

Following functional failures of a number of small pyrotechnically ac- 
tuated devices, a need was recognized for an improved- output gas gen- 
erating cartridge, as well as test methods to define performance. No 
cartridge was discovered within the space arena that had a larger output 
than the NASA Standard Initiator (NSI) with the same important fea- 
tures, such as electrical initiation reliability, safety designs, structural 
capabilities and size. Therefore, this program was initiated to develop 
and demonstrate an NSI-derived Gas Generating Cartridge (NGGC). 
The objectives of maintaining the important features of the NSI, while 
providing considerably more energy, were achieved. In addition, the 
test methods employed in this effort measured and quantified the energy 
delivered by the NGGC. This information will be useful in the appli- 
cation of the NGGC and the design of future pyrotechnically actuated 
mechanisms. 



Introduction 

Failures have occurred in several small pyro- 
technically actuated devices, which attempted to 
use the NASA Standard Initiator (NSI) as the 
sole energy source. "Small pyrotechnically ac- 
tuated devices" are defined here as those mecha- 
nisms that require 500 to 1000 inch-pounds input 
energy from a cartridge for reliable functioning. 
The NSI has been used extensively within the 
space community as both an initiator and a gas 
generating cartridge. The NSI, shown in figure 1 
and described in reference 1, is an electrically 
initiated cartridge that contains a quantity of 
pyrotechnic material. The output produced by 
this material is heat, light, gas and burning par- 
ticles, which can be used to ignite other mate- 
rials and do work. An assessment was made of 
the problems encountered in the use of the NSI 
within the NASA. A survey was conducted to 
determine if a cartridge existed or needed to be 
developed to prevent these problems. A problem 
that was immediately recognized in this survey 



was the lack of test methods to define perfor- 
mance of gas generating cartridges. Upon find- 
ing no cartridge that met the requirements set 
by this effort, an NSI-derived Gas Generating 
Cartridge (NGGC) was then developed. The ap- 
proach for this development was to modify the 
NSI to produce more gas energy, to demonstrate 
its performance through baseline firing tests, and 
to demonstrate that it could meet the NSI envi- 
ronmental qualification requirements. This sec- 
tion has been divided into subsections to de- 
scribe: (1) the failures that occurred with the 
NSI, and the lack of test methods, (2) the survey 
that was conducted to find candidate cartridges, 
(3) the objectives to develop and demonstrate an 
NGGC, and (4) the approach that was used to 
develop and demonstrate the NGGC. 

Failures and Lack of Test Methods 

The failures that have occurred in critical 
cartridge-actuated mechanisms, such as pin 
pullers (ref. 2), separation nuts and explosive 
bolts, can be attributed not only to insuffi- 



- 178 - 



WELD <REF) 

TT 



EPOXY 



ELECTRICAL PIN 




CUP. CLOSURE 
DISK .INSULATING 

DISK, INSULATING 



BRIDGEWIRE 



Figure 1. Cross sectional view of NASA Standard Initiator (NSI). 



cient input energy, but also to a lack of under- 
standing of pyrotechnic mechanisms and test- 
ing methods (ref. 3). The energy output of 
pyrotechnic gas generating cartridges is influ- 
enced by the conditions into which they are fired. 
For typical piston/cylinder configurations, these 
conditions are volume (shape and size), mass 
moved, resistance to motion (friction), and ther- 
mal absorptivity/reflectivity of the structure. 

The only existing standard for measuring car- 
tridge output performance in the field of 
pyrotechnics is the closed bomb, reference 4, 
which is inadequate for measuring energy deliv- 
ered by cartridges. The closed bomb is a fixed 
volume into which the cartridge is fired. The 
pressure produced in the volume is monitored 
with pressure transducers and the data (pressure 



versus time) are recorded. As described in ref- 
erence 5, the closed bomb provides no quanti- 
tative information that can be related to work 
performed in an actual device. 

The use of the NSI as a gas generating car- 
tridge has both advantages and limitations. The 
advantages are: (1) it is accepted in the com- 
munity with no additional environmental qualifi- 
cation required, (2) it has excellent safety fea- 
tures, 1-amp/l-watt no-fire, and electrostatic 
protection, (3) it has a demonstrated structural 
integrity and (4) it has a demonstrated reliabil- 
ity of electrical initiation and output as an ig- 
niter. The limitations are: (1) it was not de- 
signed for use as a gas generator, (2) the gas 
output produced is inconsistent in different man- 
ufacturing groups (ref. 2), and (3) it does not 



- 179 - 



provide sufficient work output for many small 
mechanisms. To overcome these problems, the 
NSI has been used with booster modules. These 
modules are sealed assemblies that contain ad- 
ditional pyrotechnic gas generating material and 
are installed into the structure of the mechanical 
device. This requires additional volume, mass 
and seals, as well as additional costs for develop- 
ment, demonstration, and qualification. 

Survey of Gas Generating Cartridges 

A survey of literature and personnel within 
NASA and Air Force space centers was con- 
ducted, using the following criteria: 

1. Provide the following important features that 
are the same as those in the NSI: 

* safety 

* reliability of electrical initiation 

* high-strength construction 

* small size 

2. Output performance greater than that of the 
NSI 

3. Long-term thermal/vacuum stability for 
space applications. 

The survey revealed the following information. 
Some cartridges do not have the NSI safety fea- 
tures. None have the NSI demonstrated reliabil- 
ity. Several use the same pyrotechnic materials 
as the NSI. Several use gas generating materials 
that are not stable under thermal/vacuum condi- 
tions. None offer sufficient advantages for further 
evaluation. 

Based on this information, a decision was made 
to develop a new, NSI-derived, Gas Generating 
Cartridge (NGGC). 

Objectives for Development and 
Demonstration of the NGGC 

The objectives of this effort were to demonstrate 
the feasibility of designing/developing an NSI- 
derived Gas Generating Cartridge (NGGC) by: 

1. Maintaining the important electrical initia- 
tion and structural reliability of the NSI 

2. Providing significantly more energy than is 
provided by the NSI 

3. Characterizing the work performance of the 
NSI and NGGC to assist in the design of 
pyrotechnically actuated devices, and 



4. Maintaining the same environmental surviv- 
ability as the NSI. 

Approach Used for Development and 
Demonstration of the NGGC 

The approach for this effort was divided 
into designing/developing, demonstrating/ 
characterizing and environmental survivability. 

Figure 2 shows the design for the NGGC, which 
utilizes the NSI body and electrical interface. 
The electrical interface is defined as the electri- 
cal pins, bridgewire, bridgewire slurry mix and 
the initial load of 40 milligrams of NSI mix. The 
remaining volume within the NSI was filled with 
a thermal/vacuum-stable gas-generating mix, in- 
cluding the additional second load above the 
ceramic cup. A joint development was con- 
ducted with the two certified NSI manufactur- 
ers, Hi-Shear Technology and Universal Propul- 
sion Company (UPCO). 

The performance of the NGGC was established 
by measuring and recording input and output 
characteristics for comparison to the same mea- 
surements in other cartridges. That is, for in- 
put electrical ignition performance tests, a direct- 
current electrical circuit was used to apply input 
electrical currents in discrete steps, measuring 
the times from current application to bridgewire 
break and to first indication of pressure. The 
output performance of the NGGC was character- 
ized with four test methods, the industry stan- 
dard and three work-measuring devices: (1) The 
Closed Bomb is the industry standard, which in- 
volves firing the cartridge into a closed, fixed vol- 
ume, measuring the pressure versus time in the 
volume, (2) The Energy Sensor, which involves 
firing the cartridge against a constant force, mea- 
suring energy as the distance stroked times the 
resistive force, (3) The Dynamic Test Device, 
which involves firing the cartridge to jettison a 
mass, determining energy by measuring veloc- 
ity of the mass to calculate l/2mv 2 and (4) The 
Pin Puller, which involves firing the cartridge to 
withdraw a pin for release of an interface, deter- 
mining energy by measuring velocity of the pin 
to calculate l/2mv 2 . 

To demonstrate the environmental survivability 
of the NGGC, the input and output performances 
of as-received (untested) units were used as per- 
formance baselines for comparison to the perfor- 
mances achieved by units that were environmen- 
tally tested. Changes in functional performance 
would indicate a sensitivity to environments. 



- 180 - 




DISK, INSULATING 



GAS-GENERATING LOAD 
85-90 MG 



Zr/KC104 
PROPELLANT 
40 MG 



Figure 2. Cross sectional view of NASA Standard Gas Generator (NSGG), showing changes to NSI configuration. 



Cartridges Tested 

Four cartridges were evaluated in this program, 
the NSI the Viking Standard Initiator (VSI), 
and two NGGC models, which were produced by- 
different manufacturers. 

NASA Standard Initiator (NSI) 

The NSI units evaluated in this program were 
manufactured by Hi-Shear Technology. Hi-Shear 
Technology utilized the same Zr/KClC>4 in both 
the NSI and their NGGC. 

Viking Standard Initiator (VSI) 

The VSI is functionally identical to the NSI 
and was manufactured by Hi-Shear Technology 
in 1972 for the Viking Program's lander on the 
surface of Mars; Since very few NSI units were 
available, the VSI functional performances were 
used to represent that produced by the NSI. 



NSI-derived Gas Generating Cartridge 

(NGGC) 

The NGGC units, manufactured by Hi-Shear 
Technology and UPCO, are the same as the NSI, 
except for the major changes shown in figure 2. 
The electrical interface from the bridgewire re- 
mained the same with the slurry mix, but the 
first press was 40 mg of Zr/KC10 4 at 10,000 psi, 
instead of the 57 mg for the NSI. The gas gen- 
erating materials were selected by each manufac- 
turer, based on a demonstrated stability against 
elevated temperature and long-term vacuum en- 
vironments. This material was pressed into 
and completely filled the charge cavity of the 
ceramic cup. An isomica insulating disc was 
bonded across the face of the cup to prevent 
an electrical path from the bridgewire through 
the pyrotechnic material to the cartridge case. 
A second increment of gas-generating material 
was pressed on top of the insulating disc to fill 
as much of the free volume as possible. The 



- 181 - 



CONSTANT 

CURRENT 

SOURCE 



CURRENT 
VOLTAGE 
MONITOR 



CARTRIDGE 




MAGNETIC 

TAPE 
RECORDER 



10 cc 

CLOSED 

BOMB 




OSCILLO- 
GRAPH 



I 



PERMA- 
NENT 
RECORD 



J 



PRES. 
(2) 
XDUCERS 



Figure 3. Cross sectional view of closed bomb and schemtic of firing and monitoring system. 



materials selected and the loading procedures 
used were the suppliers' choice and are consid- 
ered proprietary. Hi-Shear Technology loaded 
90 mg of gas generating material, while UPCO 
loaded 85, yielding a total pyrotechnic load of 
130 and 125 mg respectively, as compared to the 
114-mg load for the NSL The same electrical and 
thermal insulation discs used at the output end 
of the NSI load were also installed on the NGGC. 

Test Apparatus and Methods 

To characterize the input/output performance of 
the test cartridges, an Electrical Firing Circuit 
was used for input measurements, and four dif- 
ferent test methods were used for output mear 
surements. These output test methods were: the 
Closed Bomb, the Energy Sensor, the Dynamic 
Test Device, and the Pin Puller. All output test 
hardware was made of steel and was reusable. 

Electrical Firing Circuit 

A direct current firing circuit, shown schemat- 
ically in figure 3 and described in reference 4, 



was employed to measure the electrical igni- 
tion characteristics (function times) of cartridges 
tested. Long-duration, square-wave electrical 
pulses were applied at levels of 20, 15, 10, 5 and 
3 amperes. The input current and the pressure 
produced by the cartridges in the various out- 
put test methods were recorded on an FM mag- 
netic tape recorder with a frequency response 
that was flat to 80 Khz. Electrical initiation 
function times were measured from application 
of current to bridgewire break and from applica- 
tion of current to first indication of pressure from 
the cartridge. 

Closed Bomb 

The closed bomb, shown in figure 3 and de- 
scribed in references 3 and 4, is the industry 
standard for measuring the output of cartridges. 
The cartridge is fired into a fixed, 10-cc cylindri- 
cal volume, and the pressure produced is mea- 
sured with the pressure transducers, recorded on 
the FM tape recorder. The data collected are 
the peak pressure and the time to peak pres- 
sure. This approach has limitations, as described 



- 182 - 



in reference 5, in trying to relate the pressure 
produced to a mechanical or ignition function. 
For example, the NSI performance requirement 
in reference 1 is 650 +/— 125 psi peak pressure, 
achieved within 5 milliseconds at direct-current 
inputs of five amperes or greater. These data 
provide no quantitative information that can be 
related to work performed in an actual device. 

Energy Sensor 

The Energy Sensor, shown in figure 4 and de- 
scribed in references 4 and 5, represents an appli- 
cation where the cartridge output works against 
a constant resistive force. This resistive force is 
provided by precalibrated, crushable aluminum 
honeycomb. The strength of the honeycomb se- 
lected for this study was 500 pounds force. The 
cartridge is fired on the axis of a piston/cylinder 
as shown. The amount of work accomplished is 
obtained by multiplying the length of honeycomb 
crushed during the firing by the honeycomb's 
crush strength to yield an energy value in inch- 
pounds. 

Dynamic Test Device 

The Dynamic Test Device, shown in figure 5 
and described in references 4 and 5, represents 
the jettisoning of a mass. A one-inch diam- 
eter, one-pound, cylindrical mass is jettisoned 
through a one-inch stroke, when the o-ring clears 
the cylinder. The velocity of the mass is mea- 
sured electronically by a grounded needle on 
the mass successively contacting spaced, charged 
foils to trigger electronic pulses. These pulses 
were recorded on a magnetic tape recorder to 
measure the time interval between contact of the 
needle with the foils. Velocity was calculated by 
dividing the spacing distance (0.250 inch) by the 
time interval. The energy of the mass was calcu- 
lated as 1/2 mv?, where m is the total mass of 
the 1-pound mass and the needle. The pressure 
in the working volume was measured by a pres- 
sure transducer installed in the port as shown, 
and was recorded on the same magnetic tape 
recorder. The data collected were the energies 
delivered in inch-pounds and the peak pressures 
achieved. 

Pin Puller 

The Pin Puller, shown in figure 6 and described 
in reference 3, represents a pyrotechnic function 
with a low-mass retractor. It also presents a tor- 
tuous flow path of gases from the cartridge to 



the working piston. The cartridge's output gas, 
generated 90° from the working axis of the pis- 
ton/pin, must vent through a 0.1-inch diameter 
orifice into the working volume. Energy was ob- 
tained and calculated by measuring the velocity 
of the piston/pin, as described for the Dynamic 
Test Device. Pressure in the working volume was 
measured by a transducer installed in the sec- 
ond port, as shown. The data collected were the 
energies delivered in inch-pounds and the peak 
pressures achieved. 

Test Procedure 

The testing effort was divided into three major 
areas, as summarized in table I. This table shows 
the number of units fired in each test, as well as 



Table I. Allocation of Cartridge Test Units 
Performance Baseline Firings 



Test condition 


VSI 


NSI 


NGGC 


Hi-Shear 


UPCO 


Closed bomb 
Energy sensor 
Dynamic test device 
Pin puller 


13 
5 
8 
5 


3 

3 


6 
5 

7 
7 


5 
5 
5 
5 


Total 


31 


6 


25 


20 



Environmental Testing 



Temp, 
cycling 


Mech. 
vibr. 


Mech. 
shock 


Thermal 
shock 


NGGC 


Hi-Shear 


UPCO 


Group 1 
Group 2 
Group 3 
Group 4 


2 
3 

4 


3 

4 


4 


16 
16 
16 
16 


12 
12 
12 
13 




Total 


64 


49 



The units were visually, electrically and x-ray inspected 
before and after exposure to each environment. 

Post-Environment Firings 



Test condition 


NGGC 


Hi-Shear 


UPCO 


Closed bomb 
Energy sensor 
Dynamic test device 
Pin puller 


16 
16 
16 
16 


12 
12 
12 
13 


Total 


64 


49 



Units from the environmentally exposed groups were 
equally subdivided into each functional test group. 



Total NGGC test units: 89 69 



- 183 - 



Energy sensor 



Initiator firing block 



Cylinder p Anvil Honeycomb 

retainer 



Piston Adapter- 




Piston seal 



Figure 4. Cross sectional view of McDonnell Energy Sensor. 



Pressure transducer face 



Cartridge port 




Cyl inder 



Piston 



Sealing ring 



Figure 5. Cross sectional view of Dynamic Test Device. 



- 184 - 



Cartridge Port 



Orifice (2) 

Energy Absorbing Cup 



O-Rings (5) 




1/4" Pin 



Pressure Transducer Port- 
Figure 6. Cross sectional view of NASA Pin Puller. 



the number of units subjected to the various en- 
vironments. The Electrical Firing Circuit was 
used to collect input electrical ignition charac- 
teristics (function time) data for all units (except 
the NSI) that were fired in the Performance Base- 
line and in the Post-Environment tests. Elec- 
trical inspections were accomplished on all units 
at the start of the program with 50-volt bridge- 
to-case and 10 milliampere bridgewire resistance 
measurements. 

Performance Baseline Firings 

The performance baseline firings were conducted 
with as-received cartridges to provide a func- 
tional reference for comparisons among all 
cartridge types and to compare with post- 
environmental performance of the NGGC. Per- 
formance data included input electrical ignition 
and the four output measurements. The Elec- 
trical Firing System was used as the input fir- 
ing source for all cartridges, except the NSI. 
(NSI firings were conducted prior to the use of 
the Electrical Firing System). The cartridges 
were subdivided as equally as possible for fir- 
ings at current levels of 20, 15, 10, 5 and 
3 amperes. As an example of output test fir- 
ings described in table I, the Closed Bomb was 
used for 13 VSI, 3 NSI, 6 Hi-Shear Technology 
NGGC, and 5 UPCO NGGC units. Due to their 
scarcity, no NSI units were functioned in the En- 
ergy Sensor or Pin Puller. VSI units were used to 
supplement the data collected on the NSI, since 



their design and performance were essentially the 
same. 

Environmental Testing 

Environmental tests, duplicating the qualifica- 
tion levels and test order required for the NSI, 
were conducted on the NGGC. The NGGC units 
from each source were divided into four groups 
for testing as shown in table I. Thermal stabi- 
lization in these tests was established by thermo- 
couples attached to several cartridges; once the 
desired temperature level was indicated by the 
thermocouples, the units were soaked for at least 
15 minutes before the environmental tests began. 
The units were visually, electrically and x-ray in- 
spected before and after each environment. Elec- 
trical inspection was accomplished with 50-volt 
bridge-to-case and 10 milliampere bridgewire re- 
sistance measuapftments. 

The definition of each environmental test follows: 

Temperature cycling. The units were placed in 
a wire basket for transfer between chambers that 
were stabilized at -260 and +300°F. The follow- 
ing describes one of twenty cycles conducted: 

1. Insert units into -260°F chamber, and once 
stabilized, maintain that temperature for one 
hour 

2. Transfer units to +300°F, and once stabilized, 
maintain that temperature for one-half hour 



- 185 



2. Transfer units to +300°F, and once stabilized, 
maintain that temperature for one-half hour 

Mechanical vibration. The units were 
mounted into test blocks in a thermal chamber 
for vibration tests on all three axes. Two series of 
tests were conducted at +300 and -260°F. The 
units were conditioned at each temperature and 
the following spectrum, which produced an over- 
all G rms value of 27.5, was applied for 7.5 min- 
utes in each axis: 



Frequency (Hz) 


Level (G/Hz) 


10 - 100 
100 - 400 
400 - 2000 


0.01 to 0.8 (6 db/oct increase) 

0.8 constant 

0.8 to 0.16 (3 db/oct decrease) 



Mechanical shock. The units were mounted 
in the same test blocks used for the vibration 
tests. The units were subjected to +/— pulses on 
each axis to the following trailing edge sawtooth 
pulse: 100 G peak with an 11 ms rise and a 
1 ms decay. Tests were conducted at laboratory 
ambient conditions. 

Thermal shock. The units were placed in a 
wire basket and immersed in a container of liq- 
uid nitrogen and allowed to stabilize (no bub- 
bles). The units were then removed from the 
nitrogen and allowed to stabilize at room tem- 
perature with no protection from water conden- 
sate. The units were subjected to this process for 
five cycles, except during the fifth cycle, follow- 
ing stabilization, the units were held in the liquid 
nitrogen for 11 hours. 

Post-Environment Firings 

Following the environmental exposures, the 
NGGC units were subdivided equally and fired 
with the electrical firing circuit using the four 
test methods. That is, four units of the 16 envi- 
ronmentally tested in Group 1 were fired in each 
test method. The data collected were compared 
to the performance baseline. 

Results 

The results of the tests are presented in the same 
format as the Test Procedure section. 

Performance Baseline Firings 

The data collected for the functional performance 
baselines (input electrical function time and out- 
put tests) for each cartridge are summarized in 



the top portions of tables II through VI and fig- 
ures 7 through 10. 

Table II. Electrical Ignition Performance Data 

on Cartridges 

(Average/Standard Deviation) 





Current 




Time to 


Time to 




applied 


No. 


BW break, 


first press, 


Cartridge 


amperes 


fired 


ms 


ms 


Performance baseline (no environments) 


VSI 


20 


6 


.124/.005 


.158/.014 




15 


3 


.182/.007 


.215/.013 




10 


2 


.354 


.389 




5 


1 


1.431 


1.480 




3 


1 


121.020 


121.060 


Hi-Shear 


20 


1 


.120 


.175 


NGGC 


15 


1 


.190 


- 




10 


1 


.335 


.362 




5 


1 


1.350 


1.370 




3 


1 


12.400 


12.425 


UPCO 


20 


3 


.130/.030 


- 


NGGC 


15 


1 


.160 


.225 




10 


2 


.324 


.373 




5 


3 


1.295/.018 


1.298/.020 




3 


2 


8.355 


8.373 


Post environments 


Hi-Shear 


20 


16 


.135/.012 


.187/.025 


NGGC 


15 


12 


.184/.020 


.224/.029 




10 


12 


.329/.044 


.443/.199 




5 


12 


1.245/.050 


1.274/.063 




3 


12 


10.653/5.452 


11.315/5.789 


UPCO 


20 


12 


.121/.013 


.792/.663 


NGGC 


15 


12 


.176/.012 


.639/.481 




10 


10 


.348/.045 


.739/. 441 




5 


8 


1.238/.100 


1.327/.145 




3 


8 


7.641/4.015 


7.353/3.746 



Table III. Closed Bomb Performance Data 
on Test Cartridges 

(Average/Standard Deviation) 



Cartridge 


No. 

fired 


Time to 

peak pressure, 

ms 


Peak 

pressure, 

psi 


Performance baselii 


ae (no environmei 


its) 


VSI 

NSI 

Hi-Shear NGGC 

UPCO NGGC 


13 
3 
6 
5 


.09/.05 
.23/.06 

1.15/.34 
.48/.13 


675/81 

660/53 

1083/41 

1120/58 


Post environments 


Hi-Shear NGGC 
UPCO NGGC 


16 
12 


1.11/.30 
.28/.07 


1076/25 
1250/47 



- 186 - 



Table IV. Energy Sensor Performance Data on 
Test Cartridges 

(Average/Standard Deviation) 



Cartridge 


No. 
fired 


Energy delivered, 
inch-pounds 


Performance baseline (no environment) 


VSI 

Hi-Shear NGGC 

UPCO NGGC 


5 
5 
5 


466/21 
815/99 
812/90 


Post environments 


Hi-Shear NGGC 
UPCO NGGC, 


16 
12 


869/80 
927/58 



V8I 

HI-SHEAfl, NO ENV 
HhSHEAfl, POST ENV 
UPCO, NO ENV 
UPCO, POST ENV 




S 10 10 

Current Applied, amperes 



Table V. Dynamic Test Device Performance Data 
on Test Cartridges 

(Average/Standard Deviation) 



Cartridge 


No. 
fired 


Energy 

delivered, 

inch-pounds 


Peak 

pressure, 

psi 


Performance basel 


ine (no environ: 


ment) 


VSI 

NSI 

Hi-Shear NGGC 

UPCO NGGC 


8 
3 

7 
5 


337/64 
351/15 
785/66 
756/74 


5580/940 
5540/755 
4983/993 
9408/2002 


Post environments 


Hi-Shear NGGC 
UPCO NGGC 


16 
12 


667/45 

777/50 


6953/1866 
8337/1980 



Table VI. Pin Puller Performance Data on 
Test Cartridges 

(Average/Standard Deviation) 



Cartridge 


No. 
fired 


Energy 

delivered, 

inch-pounds 


Peak 

pressure, 

psi 


Performance baseline (no environment) 


VSI 

Hi-Shear NGGC 

UPCO NGGC 


5 
7 
5 


154/20 
450/36 
526/39 


7056/143 

12313/797 
16392/1514 


Post environments 


Hi-Shear NGGC 
UPCO NGGC 


16 
13 


462/21 
514/17 


11720/440 
15532/680 



V8I 

-#- HI-SHEAR, NO ENV 
-9- HI-SHEAR, POST ENV 
-*~ UPCO, NO ENV 
~9- UPCO. POST ENV 



0.1 x 





5 10 IS 

Current Applied, amperes 



Figure 7. Plots of current applied versus time to bridge- 
wire break (top) and to first pressure (bottom). 



Electrical ignition performance. The elec- 
trical ignition performance baselines (no envi- 
ronments) for the VSI and NGGC are shown in 
table II and figure 7. As mentioned earlier, no 
NSI data were collected. Each of the data points 
on the plots is the averaged value of the func- 
tion times for each cartridge type at each current 
level. Very little difference in performance could 
be detected among any of the cartridge groups. 

Closed bomb. The closed bomb performance 
baseline data are shown in table III. Typical 
traces for each cartridge type are shown in fig- 
ure 8. The times to peak pressure for the VSI 
and NSI are smallest. The average time to peak 
pressure for the Hi-Shear Technology NGGC is 
considerably longer than for the UPCO units 
(1.15 versus 0.48 ms). The NGGC peak pressures 
achieved are comparable (1083 and 1120 psi). 

Energy sensor. The Energy Sensor perfor- 
mance baseline data are shown in table IV. The 



- 187 - 



1600 - 




Hi-Shear t?CGC 
UPCO NGCC 



Figure 8. Typical pressure traces produced by the NSI, VSI and UPCO, and Hi-Shear NGGC in the closed bomb. 




Figure 9. Typical pressure traces produced by the VSI and Hi-Shear and UPCO NGGC in the dynamic test device. 



- 188 - 



16000 - 



-UPCO NCGC 



12000 



8000 



4000- 




Time, millisecond 
Figure 10. Typical pressure traces produced by the VSI and Hi-Shear and UPCO NGGC in the pin puller. 



NGGC performances are comparable (815 and 
812 inch-pounds) and nearly twice that of the 
VSI (466 inch-pounds). 

Dynamic test device. The Dynamic Test De- 
vice performance baseline data are shown in ta- 
ble V. The performance of the VSI and NSI 
are comparable (337 and 351 inch-pounds), as 
are the two NGGC models to each other (785 
and 756 inch-pounds). The NGGC performance 
is over twice that of the VSI and NSI. Figure 9 
shows typical pressure traces from each cartridge 
type. The peak pressure for the UPCO NGGC 
is appreciably higher and more dynamic than the 
Hi-Shear Technology units. 

Pin puller. The Pin Puller baseline data 
are shown in table VI. The performance of the 
NGGC models (450 and 526 inch-pounds) are 
three times that produced by the VSI (154 inch- 
pounds). Figure 10 shows distinctively different 
pressure traces from each cartridge type. 

Environmental Testing 

The environmental testing was completed with 
no evidence of physical damage through visual, 
x-ray and electrical inspections. 



Post-Environment Firings 

The data collected for the post-environment 
functional performance tests (input electrical 
function time and output tests) for each cartridge 
are summarized in the lower portions of tables II 
through VI. 

Electrical ignition performance. The elec- 
trical ignition data are shown in table II and fig- 
ure 7. The only change between pre- and post- 
environment performance was a small increase in 
times to first indication of pressure in the UPCO 
NGGC units. All values were within the 5 mil- 
lisecond delay times at 5 amperes or greater, al- 
lowed by the NSI specification, reference 3. 

Closed bomb performance. The data fol- 
lowing environmental exposure are shown in 
table III. No appreciable change in performance 
was observed. 

Energy sensor. The Energy Sensor baseline 
data are shown in table IV. The apparent in- 
crease in energy delivered (54 and 115 inch- 
pounds) by the NGGC models following envi- 
ronmental exposures is insignificant, considering 



- 189 - 



the standard deviations of the pre- and post- 
environments data. These standard deviations 
total 179 and 148 inch-pounds for the respective 
NGGC models, which could include this range of 
data. 

Dynamic test device. The Dynamic Test De- 
vice data are shown in table V. No significant 
change in performance was observed following en- 
vironments, again considering the standard devi- 
ations. 

Pin puller. The Pin Puller data are shown 
in table VI. No change was detected following 
environments. 

Conclusions 

All objectives of this effort were met, which 
should allow for immediate consideration for the 
application of the NGGC. The NGGC was man- 
ufactured using the same body and electrical 
interface as the NSI. The electrical initiation 
characteristics are the same as the NSI. A slight 
delay was observed in the time to first pressure 
for the UPCO NGGC following environmental 
exposures. This delay, caused by a decrease in 
thermal transfer from the bridgewire, is accept- 
able, since it is well within the NSI performance 
specification. The cartridge functional evalua- 
tions used in this effort clearly show that out- 
put working energy is affected by the configura- 
tion in which it is used. The Energy Sensor and 
Dynamic Test Device measured the most energy 
delivered by the NGGC, about 800 inch-pounds, 
while the Pin Puller was much less efficient, de- 
livering only about 500 inch-pounds. Although 
the two NGGC manufacturers selected different 
thermal/vacuum-stable gas generating materials, 
as evidenced by the different pressure traces ob- 
served, the output performance of the two models 
was essentially the same in each of the four test 
methods. Under the assumption that the NSI 
produces the same output performance as the 
VSI, the NGGC produces approximately twice 
the output of the NSI/VSI in the Energy Sensor 
and the Dynamic Test Device, and three times 
that of the NSI/VSI in the Pin Puller. No signif- 
icant change in output performance was observed 
following exposure of the NGGC to the rigorous 
thermal/mechanical environmental requirements 
for the NSI. 

A work-producing cartridge has been developed 
with the key attributes of the NSI. Designers 



of pyrotechnic mechanisms now have a cartridge 
that is defined in terms of work, and they can 
relate the test configurations and energy deliv- 
eries documented in this report to their design 
requirements. This cartridge performance can 
meet the requirements of a substantial portion 
of small aerospace pyrotechnic devices (including 
many applications where the NSI with booster 
modules are now employed), such as pin pullers, 
nuts, valves and cutters. 

A word of caution is warranted. The suc- 
cessful completion of this developmental effort 
does not "qualify" the NGGC for any appli- 
cation. Users must conduct a developmental 
effort, including demonstrating functional mar- 
gins, environmental qualification, and system in- 
tegration/operation demonstrations for devices 
in which the NGGC is to be used. 

The acquisition of the NGGC should be based 
on performance, as measured by at least one of 
the energy measuring devices described in this 
report, the Energy Sensor, the Dynamic Test 
Device or the Pin Puller. 

References 

1 . Design and Performance Specification for 
NSI-1 (NASA Standard Initiator-1), 
SKB26100066, January 3, 1990. 

2. Bement, Laurence J. and Schimmel, 
Morry L.: "Determination of Pyrotechnic 
Functional Margin." Presented at the 1991 
SAFE Symposium, November 11-14, 1991, 
Las Vegas, Nevada. Also presented at the 
First NASA Aerospace Pyrotechnic Systems 
Workshop, June 9-10, 1992, NASA Lyndon B. 
Johnson Space Center, Houston, TX. 

3. Bement, Laurence J.: "Pyrotechnic System 
Failures: Causes and Prevention." NASA 
TM 100633, June 1988. 

4. Bement, Laurence J.: "Monitoring of Explo- 
sive/Pyrotechnic Performance." Presented 
at the Seventh Symposium on Explosives 
and Pyrotechnics, Philadelphia, Pennsylva- 
nia, September 8-9, 1971. 

5. Bement, Laurence J. and Schimmel, 
Morry L.: "Cartridge Output Testing: Meth- 
ods to Overcome Closed-Bomb Shortcom- 
ings." Presented at the 1990 SAFE Sympo- 
sium, San Antonio, Texas, December 11-13, 
1990. 



- 190 - 



Coin 



t 







DEVELOPMENT OF THE TOGGLE DEPLOYMENT MECHANISM 

Christopher W. Brown 
NASA Lyndon B. Johnson Space Center, Houston, Tx 



Abstract 

The Toggle Deployment Mechanism 
(TDM) is a two fault tolerant, single 
point, low shock pyro/mechanical 
releasing device. Many forms of 
releasing are single fault tolerant 
and involve breaking of primary 
structure. Other releasing 

mechanisms, that do not break 
primary structure, are only 
pyrotechnically redundant and not 
mechanically redundant. The TDM 
contains 3 independent pyro 
actuators, and only one of the 3 is 
required for release. 

The 2 separating members in the 
TDM are held together by a toggle 
that is a cylindrical stem with a 
larger diameter spherical shape on 
the top and flares out in a conical 
shape on the bottom. The spherical 
end of the toggle sits in a socket 
with the top assembly and the 
bottom is held down by 3 pins or 
hooks equally spaced around the 
conical shaped end. 

Each of the TDM's 3 independent 
actuators shares a third of the 
separating load and does not 
require as much pyrotechnic 
energy as many single fault 
tolerant actuators. Other single 
separating actuators, i.e., 
separating nuts or pin pullers, have 
the pyrotechnic energy releasing 
the entire preload holding the 
separating members together. 

Two types of TDM's, described in this 
paper, release the toggle with pin 
pullers, and the third TDM releases 
the toggle with hooks. Each design 
has different advantages and 
disadvantages. This paper describes 



the TDM's construction and testing 
up to the summer of 1993. 

Introduction 

Numerous aerospace programs have 
been a need for low shock, single 
point separators that are multi-fault 
tolerant and can separate without 
breaking primary structure. In late 
1987, such requirements were 
applied on an Orbiter Disconnect 
Assembly that is part of the 
Stabilized Payload Deployment 
System (SPDS). Two different types 
of TDM's were developed. They 
differed by way of solving an initial 
design problem. Both TDM's had 3 
pin pullers each and were 
vulnerable to unexpected tensile 
load spikes creating plastic 
deformation. If the pins were bent, 
they could not retract which would 
lead to a separation failure with 
little to no preload holding the 2 
structures together. 

In late 1988 a request for a TDM in a 
different envelope shape was 
considered, and a third concept, the 
TDM20KS, was designed. This TDM 
was met the envelope constraints 
and was designed to function with 
the inner parts deformed from 
tensile load spikes. The TDM20KS 
application dissolved, but the 
development tests continued. 

Most of the tests performed 
considered the 2 fault tolerant case 
by using only one of the 3 
separation members. 

The First Toggle Deployment 

Mechanisms 



- 191 - 



The original SPDS requirements 
were translated into the first pin- 
puller configured TDM and resulted 
in the United States Patent 
4,836,081.!. The original concept, 
shown in figure 1, had a problem 
with the preload forcing the pins 
back and causing the shear pins to 
function as primary structure. The 
less the angle "a" from the 
horizontal on the toggle, the less 
load there is pushing the pins back. 
However, the less angle "a", the 
harder it is for the toggle to swing 
away from the unretracted pins. If 
angle "a" was zero, there would be 
no moment on the toggle, created 
by the 2 unretracted pins, for the 
toggle to swing away from. 

The first NASA-JSC pin-puller 
configured TDM, seen in figure 2, 
has the axis of the pins parallel to 
the conical surface of the toggle. 
The tension of the toggle pushes the 
conical surface perpendicular to 
the pin ! s axis and does not act on 
the shear pins. 

Preloading the toggle was 
completed by unscrewing the 
preload collar that pulls the toggle 
up. To prevent twisting of the 
toggle wile preloading, a tool was 
placed in the socket holes to hold 
the toggle and socket from turning. 
This TDM was designed to hold and 
release a preload of 1000 pounds. 

Testing was performed with 
pneumatics and NASA Standard 
Initiators (NSI). With pneumatics, a 
static pressure of about 500 psi was 
needed to release the 1000 pound 
preload. Dynamic pressure 

releasing was performed by 
opening a solenoid valve into the 
NSI port. Pluming orifices and 
other dynamics required a higher 
static pressure behind the solenoid 
valve for separation. Figure 3 
shows pressure versus preload with 
two types of piston/pin coatings. 



Friction coefficients of the TDM 
parts play an important role in 
releasing energy. Figure 3 also 
shows a difference between the 
original piston finish and a Teflon 
impregnated nickel plate finish 
called NEDOX from General 
Magniplate Corp.. The NEDOX finish 
shows a consistent improvement in 
releasing energy. 

The second TDM concept had a 
different design to avoid 
transferring the preload into the 
axis of the pins. This resulted in 
development of the double swivel 
toggle and lead to United Stated 
Patent 4,864,910. 2 . As seen in Figure 
4, this TDM had the axis of the pins 
running perpendicular the axis of 
the toggle. When one pin was 
retracted, the bottom of the toggle 
could swivel down and clear itself 
from the 2 unretracted pins. This 
TDM was successfully developed and 
qualified to meet SPDS 
requirements. 

Toggle/Hook D eployment 
Mechanism TDM20KS 

The TDM20KS was designed to 
release a preload up to 20,000 
pounds even if some internal parts 
were plastically deformed. Figures 
5 and 6 show the fastened and 
released configurations of the 
internal parts. In figure 5, the 
toggle is held down with 3 hooks 
that pivot in the body. Each hook is 
held from pivoting by a piston. As 
seen in figure 6, the piston has 
moved up and the toggle is released 
when a hook is free to swing back 
into the void of the piston. 3, 

Two TDM20KS's were made with 
differences in the angle of 
toggle/hook contact. One TDM20KS 
had a 45 degree angle of contact, 
and the other assembly had a 30 
degree angle contact from the 



- 192 - 



horizontal. Both toggle stems had 
full strain gauge bridges applied 
inside holes going through their 
axes. Each piston port had 2 NSI 
ports. Most parts of the TDM20KS 
were plated or coated with low 
friction surfaces. 

Tnyr?nK<s TVvPlnpment Test 

Three sets of development tests 
were completed. After initial 
testing, the next 2 development tests 
were performed with design 
changes that were needed during 
the initial testing. The initial 
development tests were performed 
with hydraulics and with NSI's, but 
the first goal was applying the 
preload. 4 . With 2 NSI ports per 
cylinder, pressure monitoring was 
easily accomplished. 

TDM20KS Tnitial Dev elopment 
Testing 

Preloading was similar to the 
original TDM by way of pulling the 
socket up when unscrewing the 
preload collar (otherwise known as 
a preload bolt). Figure 7 shows the 
setup used to get the maximum 
preload. A tension machine would 
pull the socket up by stretching the 
toggle, and the preload bolt was 
unscrewed until it was snug with 
the socket. The bolts connecting 
the tension machine to the socket 
had a 17,000 pound maximum limit, 
and the TDM20KS assembly would 
settle down to about 12,000 pounds 
preload. A future design was able to 
obtain a 20,000 pound preload. 

Releasing the preload with 
hydraulic pressure in one cylinder 
showed the difference between the 
45 degree TDM and the 30 degree 
TDM. Figure 8 shows 3 different 
releasings at 3 different preloads 
for the 45 degree TDM, and figure 9 
shows the same tests with the 30 
degree TDM. As expected, there was 



less pressure needed to release the 
30 degree TDM due to less force 
between the hooks and pistons. 
What was unexpected is the 
inconsistency of releasing pressure 
as a function of preload. 

NSI firing showed similar 
pressure/preload results after the 
TDM's were exposed to vibration, 
shock, and thermal environments. 
Ambient firings revealed a design 
problem in the piston stops. The 
pistons traveled too far and leaving 
the bottom of the piston voids 
pushing the hook back up which 
prevented toggle releasing. All 
toggle releases, with the original 
piston stops, were performed with 
one NSI per piston. The second 
phase of development testing, with 
2 NSFs per piston, showed some 
success in an improved design. 

Figures 10 and 11 show pressure 
and preload curves in the 30 and 45 
degree TDM chambers during 275 
degree F. firings. The TDM's were 
successfully fired in -90 degree F. 
environment. Figure 12 shows the 
data on the 45 degree TDM cold 
firing. 

Both hydraulic and NSI tests showed 
a preload increase when 
separating. This is belived to be 
caused by the piston bending in 
toward the hook while traveling up. 

A plastic deformation test was 
performed with the 45 degree TDM 
while the 30 degree TDM was saved 
for testing improved piston stops 
and preloading mechanisms. 

The 45 degree TDM was first 
preloaded, and the separating 
members were tensioned to 29,000 
pounds. There was a .040 inch gap 
in the separation plane, but no 
separation was noticed while the 
tension was within the preload. The 
29,000 pound load was released, and 
the remaining preload in the TDM 



- 193 - 



was unknown due to strain gauge 
damage in the toggle stem. Figure 
13 shows the average permanent 
change in the deformed parts. The 
45 degree TDM was preloaded and 
put back into tension. The toggle 
stem finally broke at 34,000 pounds 
tension. Besides the toggle, the 
internal parts were deformed a 
little more and were still able to 
function in the TDM body. 

TDM20KS Post Development Tests 

The 2 post development tests 
evaluated a redesigned piston stop 
and a new type of preloading 
mechanism. 

Piston Stop Redesign 

The initial NSI firing tests revealed 
that the piston stops were not 
stopping the piston with the 
actuation of one NSI. The existing 
stops, made of AL7075-T6, were set to 
start taper locking the piston tops at 
a position too high for the piston to 
settle in the correct position. In 
addition, the original piston stops 
were cracking at the sides where 
the bolts held them down. The 
original piston stops were lowered 
and extra side supports were bolted 
down, but these created assembly 
problems that led to the new piston 
stop. 

Besides making the new piston stops 
out of 15-5 CRES, the design was 
beefed-up on the outside, and the 
interior dimensions had tighter 
tolerances. Figure 14 shows the 
difference in the piston stops. 

Six firings of the TDM were 
conducted at various temperatures, 
preloads, and number of NSI'sA 
Most tests were performed with 2 
NSI's per piston unlike the initial 
development tests. Tests proved the 
piston stops did not deform, but the 
original 1/4-20UNC-3A bolts, 



holding the stops, elongated and 
were bent. Figure 15 shows how the 
deformed bolts allowed the piston to 
travel too far, pushing the hook 
back up, and wedging the piston 
between the hook and cylinder. 
Testing with bolts made of carbon 
steel, instead of 300 series SST, also 
resulted in stretched bolts but not as 
deformed as the older ones. The 
bottom surface of the piston void 
was lowered 0.15 inchs to give room 
for successful releasing with 2 NSI's 
per piston. 

Top Member Redesign 

The last of the TDM20KS 
development solved the preloading 
problem by completely redesigning 
the top preloading members. As 
noticed in earlier TDM preloading, 
one set of threads could apply a 
certain tension regardless to the 
diameter. Also, the tension 
machine would have to load the 
toggle about 1.4 times the desired 
preload to get what the TDM would 
settle down to. The theory behind 
the new top TDM section is to design 
in as many sets of threads as 
possible, and the tensions from all 
threads would sum up to a desired 
preload. 

Figure 16 shows the major parts 
used in the new top members 
designed to get up to 20,000 pounds 
preload without using a tensile 
machine. The only existing part 
used is the socket. It sits in a 
tensioner that contains 18 sets of 
3/8"-24 threaded holes surrounding 
the socket. Eighteen hex cap screws 
were threaded into the tensioner 
and sit in the base plate. This base 
plate has 18 counter bores with the 
same bolt circle as the tensioner 
threads. Each base plate counter 
bore has 2 SST washers with a brass 
washer between. These washers act 
as thrust bearings for the bolts. 
Tightening the bolts in a star 



- 194 - 



pattern would separate the 
tensioner from the base plate and 
create tension on the toggle. 

Testing of the new top member and 
releasing over 20,000 pound preload 
with one NSI was successful. 6 . 
Each bolt was torqued at 20 in. lb. 
increments, in a star pattern, until 
the toggle's strain gauge failed at 
18,000 pounds. Preload, as a 
function of torque, was continued 
until 20,000 pounds was estimated. 
The maximum torque/preload tested 
was 200 in.lb. on each bolt and 
24,000 pounds preload. At ambient 
conditions, one NSI was still able to 
release the 24,000 pound preload. 

Conclusions 

Each of the three toggle deployment 
mechanism concepts tested 
successfully. Other designs were 
considered which had different 
hook shapes and pivot locations. 
Besides pivoting hooks, sliding 
members can also apply to fit 
envelope constraints. 

References 

1. Patent Number 4,836,081 Jun. 
6,1989 "TOGGLE RELEASE". 
Inventors: Thomas J. Graves; 
Robert A. Yang; Christopher W. 
Brown. Assignee: The Unites 
States of America as represented 
by the Administrator of the 
National Aeronautics and Space 
Administration, Washington, D.C. 

2. Patent Number 4,864,910 
"DOUBLE SWIVEL TOGGLE 
RELEASE". Inventors: Guy L. 
King; William C. Schneider. 
Assignee: The Unites States of 
America as represented by the 
Administrator of the National 
Aeronautics and Space 
Administration, Washington, D.C. 

3. NASA Tech Briefs Vol. 15 No. 11 
P. 71 "Redundant Toggle/Hook 
Release Mechanism" Nov. 1991. 



6. 



NASA JSC Thermochemical Test 
Area No. JSC 26486 "Internal 
Note For the Toggle Deployment 
Mechanism (TDM)" April 1993. 
NASA JSC Thermochemical Test 
Area No. JSC 25964 "Internal 
Note For Toggle Deployment 
Mechanism Piston Stop Test" 
July 1992. 

NASA JSC Thermochemical Test 
Area Task History 2P809 " August 
1993. 



- 195 - 




Shear Pin 



^- F - Preload/Tan a 



Figure I 

Original Toggle Deployment Mechanism Concept. 



- 196 - 



Preload Collar/Bolt 



Top Plate 



Piston/Pin 



Collar 



Shear Pin 



Toggle 




«•- Separation Plane 



O-Ring 



O-Ring 



Body 



Figure 2 

Toggle Deployment Mechanism, Pin-Puller Configuration. 



- 197 - 



CO 

oo 



u 

Cl-I 

C 
O 








c 


a 


o 


CO 


3 


D-, 


CO 

to 


U 


cu 




u. 


6 


a 


OJ 


4-> 


n 


U 


Q 


cu 




1-4 




«-J 




o 




c 




CO 




<V 




O 




Q 



2000 



1000 



1600 



MOO 



1200 - 



1000 



ooo 



600 



400 



:oo 









— 


1 










— 




















— 























































} 




































i 






































*■"/ 


1 


— 


— 




1 

i 















































— 


— - 




— 






























































— 


























V 




— 


— 






— 






.._. 















_J 




/ 


.__ 


— 



































■■ 'J* 


/ 






. -> 


A 
























^ 


^ 


— 


— 








/. — 













^ 


s 


— - 




— 


— 








s 


f 


<_ 






















— 
























S 




























S 














— 






















i 




















i 



-o- NEDOX(psi) 

-+- 16 micro (p3i) 



1 00 200 300 400 500 600 700 000 900 1 000 



Load (lbs) 



Figure 3 

Pressure pulse v.s. preload for releasing the TDM, 



Preloading Nut 



Two Part Toggle 



Socket 




Bottom Swivel 



Separation Plane 



Piston/Pin 



Figure 4 

Double Swivel Toggle Release. 



- 199 - 



ST0P(3-TYP) 
HOOK (3-TYP) 



PISTON (3-TYP) 



PYRO PORT (3-TYP) 




TOP CAP 



SOCKET 



PRELOAD BOLT 
LOCK PIN 



TOGGLE 



PISTON (3-TYP) 



HOOK (3-TYP) 



SECTION A-A 



Fi gure 5 

Toggle/Hook Deployment Mechanism TDM20KS before separation. 



- 200 - 



TOP CAP 



STOPO-TYP) 




PISTON (3-TYP) 



PYRO PORT (3-TYP) 




SECTION B-B 



PISTON (3-TYP) 



HOOK (3-TYP) 



ELgUI£J& 

Toggle/Hook Deployment Mechanism TDM20KS during deployment. 
Separation plane is at section B-B. 



- 201 - 




TENSION 
MACHINE 



PRELOAD 
PLATE 

SOCKET 
ADAPTOR 

SOCKET 
BOLTS 



SOCKET 

PRELOAD 
BOLT 

TOP PLATE 



THREAD LOCK 
HOLE 



TDM 



TENSION 
TOOL 



TENSION 
MACHINE 



Fi gure 7 

Preloading the TDM20KS. 



- 202 - 



TOGGLE PRELOAD 




lbf and psig 
12,000 

10,000 

8,000 

6,000 



4,000 



2,000 



HYDRAULIC PRESSURE 



Figure 8 

Hydraulic release tests for the 45 degree TDM. Load cell and toggle gauge lines 
were staggered due to pen locations on the strip chart recorder. 



- 203 - 



I 

O 
I 



TOGGLE PRELOAD 




lbf and psig 
12,000 

10.000 

8,000 
6,000 
4,000 
2,000 



HYDRAULIC PRESSURE 



Fi gure 9 

Hydraulic release tests for the 30 degree TDM. Load cell and toggle gauge lines 
were staggered due to pen locations on the strip chart recorder. 



12x10 



10- 



PRELOAD RELEASED 280 uS 
AITER PRESSURE RJSE 



-PRELOAD INCREASED 
252 lbf. DURING RELEASING 



i 

IV, 

o 
I 



LEGEND: 

NSI PRESSURE 

TOGGLE PRELOAD 




800nS 



Fi gure 10 

Forty-five degree TDM pyrotechnic release at 275 degrees F. Toggle preload 
values are not compensated for thermal effects on toggle gauges. 



10x10 - 



8 



o. 





(/> 






2 


6 


1 


a 




n: 


■a 




o 

en 


o 

r— « 

4> 




i 


fc. 


4 



2- 



-PRELOAD INCREASnD 
179 Ibf. DURING RELEASING 



PRELOAD RELEASED 340uS 
AFffiR PRESSURE RISE 



2.0 



-PEAK PRESSURE 5440 psi \ 
1 IOiiS AFTER RISE 



1^ 
2.2 



""T~ 
2.4 



2.6 



LEGEND: 

NSI PRESSURE 

TOGGLE PRELOAD 




2.8 



3.0mS 



Figure 11 

Thirty degree TDM pyrotechnic release at 275 degrees F. Toggle preload values 
are not compensated for thermal effects on toggle gauges. 



14x10 - 



12- 





CO 




M 




GO 




^ 




•a 




s 


1 


13 


ro 


O 


O 
^4 





10- 



8 - 



6- 



4 - 



2- 



PEAK PRESSURE 3896 psi 
90uS AFTER RISE 



1^ 
0.2 



0.4 



0.6 



-PRELOAD RELEASED 410uS 
AFTER PRESSURE RISE 



LEGEND: 

NSI PRESSURE 

TOGGLE PRELOAD 




0.0 



1.0mS 



Fi gure 12 

Forty-five degree TDM pyrotechnic release at -90 degrees F. Toggle preload 
values are not compensated for thermal effects on toggle gauges. 























r 

s. 


y 




1 


i 


1 


i 



.002 SHORTER 



f i \ 



.001 WIDER 




PISTONS 



PISTONS 



.003 OUT OF ROUND -i 




007 TALLER 



HOOKS 



Figure 13 

Dimensional analysis of the deformed 45 degree TDM parts after a 
29,000 lbf structural load test. Dimensions are averages of all parts and in inches. 



- 208 - 








I 







< I 



Original Piston Stop 
AL 7075-T6 



Rpdpsigned Piston Stop 
15-5PHCRES 



Figure 14 

TDM20KS Piston Stops. 



- 209 - 




Fi gure 15 

TDM20KS Piston/Hook interference after excessive piston travel. 



- 210 - 



Bolts 



Socket 



Tensioner 




Base Plate 



Separation Plane 



washers 



Toggle 



Figure 16 

TDM20KS Redesigned Top Assembly 



- 211 " 



SI*-** 



THE ORDNANCE TRANSFER INTERRUPTER. A NEW TYPE OF S&A DEVICE 

John T. Greenslade, Senior Staff Engineer 
Pacific Scientific Company, Energy Dynamics Division 



ABSTRACT 

A discussion is given in this paper of a new 
approach to the Safing and Arming of 
aerospace ordnance systems interconnected 
by detonation transfer lines, in which the 
conventional type of S&A device normally 
used for this purpose is replaced by a 
relatively simple electro-mechanical 
switching device, referred to as an 
"Interrupter. " In this approach the 
Interrupter, which is interposed in the 
transfer line between the system initiator and 
output device, is completely passive in that 
it contains no pyrotechnic devices or 
materials. Being passive (therefore, non- 
initiating), the Interrupter is much less 
hazardous to handle and install, as well as 
being significantly less complex and costly 
than conventional S&A devices containing 
EEDs and explosive leads. 

Details are presented relative to the design, 
development and qualification, by PS/EDD, 
of an ordnance transfer Interrupter intended 
for use on a commercial launch missile. 
This device, which is capable of 
simultaneously "switching" multiple 
detonation transfer lines, incorporates a rod- 
type rotary barrier with independent 
transverse apertures for each transfer line. 
Barrier actuation is bi-modal, i.e., the 
barrier can be driven from safe-to-arm or 
from arm-to-safe positions by independent 
electro-mechanical actuators. The 

Interrupter described also features visual and 
remote status monitoring provisions and, in 
common with range-approved conventional 
S&A devices, a pre-flight safety locking 
mechanism functioned by a removable safing 
key. 



Successful development of the Interrupter 
required the resolution of such problems as 
ensuring reliable detonation propagation 
(between opposed booster tips in the transfer 
lines) across unusually large airgaps within 
the barrier apertures, and the damping of the 
barrier drive train to prevent inadvertent 
actuation during vibration and shock 
extremes. The latter problem was solved by 
the incorporation of in-line pneumatic 
dampers in each of the barrier drive-trains. 

INTRODUCTION 

For safety reasons, all ordnance systems, 
regardless of their level of complexity, must 
be capable of being maintained in an 
inoperative "Safe" state, prior to when they 
are required to function. At the simplest 
level , the Safing function could be 
exemplified by the switching of a firing 
circuit of a single electro-explosive device 
(EED). In most missile and spacecraft 
ordnance systems, the Safing function, and 
its converse, the Arming function are 
effected by a specifically designed, and 
often complex, Safe/ Arm Device, or SAD. 
A brief discussion relative to the 
conventional usage and technology of these 
devices will provide an appropriate 
introduction to the subject of this paper. 

The SADs employed in missile ordnance 
systems have generally been classifiable in 
either of two broad categories, namely, the 
"Command" type and the "Inertial" type. 
SADs of the former type are "commanded" 
to change their state from Safe to Arm (and 
reverse, in some cases) by the input of 
electrical signals generated by a remote 
controller. On the other hand, the Safe to 
Arm transition is achieved automatically 



- Z13 - 



PAGE 



JUL 



.V liL, 



with the "InertiaT type SAD, when it is 
subjected to a specific level of vehicle 
acceleration for a specific minimum 
duration. The arming mechanisms in the 
Command type SADs have most commonly 
been electro-mechanical, although electronic 
devices have become competitive, and laser 
based Command type SADs have started to 
make an appearance. The Inertia! SADs are 
armed by inertia forces, resulting from the 
missile acceleration, acting on a spring- 
loaded "set-back" weight. With these 
devices, the movement of the set-back 
weight is controlled by an escapement 
mechanism, similar in principle to the 
escapements used in clocks. 

Figure 1 tabulates some of the more 
common types of SAD which have been 
used in missile applications. 

Conventional SADs, Command and Inertial, 
are more than just a system safing and 
arming mechanism; they are also the 
ordnance system initiator. For that purpose, 
they incorporate electro-explosive devices, 
usually detonators, and very often explosive 
transfer components such as "leads," or 
confined detonating cords. This, of course, 
increases the hazard potential associated 
with pre-flight testing and installation of 



TYPE 


COMBINATIONS OF: 


ENABLE 


ARM 


INTERRUPT 


COMMAND 


LANYARD 
SOLENOID 

GAS 

PRESSURE 


ROTARY 
SOLENOID 

LINEAR 
SOLENOID 

SPRING 

GAS PRESSURE 


ROTARY 
SOLENOID 

LINEAR 
BARRIER 

MOVABLE EDO 


INERTIAL 


AS ABOVE 


INERTIA 


AS ABOVE 


OTHER 


ELECTRONIC 


ELECTRONIC 


RELAY 



Figure 1: 
SADs Used in Missile Applications 



conventional SADs. As such, it is one of 
the factors which led to the concept of the 
Interrupter, which contains no pyrotechnic 
or explosive components, as a replacement 
for conventional Command SADs in systems 
employing linear ordnance transfer lines 
such as shielded Mild Detonating Cord 
(SMDC), or Confined Detonating Fuze 
(CDF). 

In the Safe mode, conventional Command 
type SADs must either physically hold their 
internal detonators out of alignment with 
their output ports, or, interpose a barrier 
between the detonators and ports. They 
must also disable the firing circuits to the 
electro-explosive detonators, and impose a 
safety shunt across those circuits. 
Conversely, when in the Armed mode, the 
Command SADs must align their detonators 
with the output ports (or remove the 
barrier), and they must complete the firing 
circuits to the detonators. In comparison, 
the Interrupter, having no internal EEDs, is 
not involved in disabling or enabling the 
firing circuits to the system initiators. This 
considerably simplifies its internal circuitry 
and switching requirements. The Interrupter 
would , as the name implies , have to 
interrupt the system firing train, in this case 
by a removable barrier. The functional 
requirements for a typical conventional 
Command SAD are compared, in Figure 2 
with those for the Interrupter. 

A number of the design requirements for 
SADs have been dictated by the 
Government' s Missile Test Ranges, 
predicated primarily on safety 
considerations. Such requirements relate, 
for instance, to the minimum amount of 
firing train misalignment needed, the 
integrity of status monitoring provisions, 
and the "hand-safe" capability of the device, 
i.e., its ability to remain intact in the event 
of an inadvertent detonation of its EEDs. 



- 2m - 



MODE 


FUNCTION 


CONVENTIONAL 
SAD 


INTERRUPTER 


SAFE 


PHYSICALLY BLOCK FIRING TRAIN 

DISRUPT FIRING CIRCUIT 

SHUNT FIRING CIRCUIT INDICATOR 


X 
X 
X 


X 


ARM 


PHYSICALLY ALIGN FIRING TRAIN 
COMPLETE FIRING CIRCUIT 
REMOVE INITIATOR SHUNT 


X 
X 
X 


X 


MONITOR 


PROVIDE INDICATION OF "SAFE" 
PROVIDE INDICATION OF "ARMED" 


X 
X 


X 
X 


SAFING 


MANUAL SAFING FROM ANY POSITION 

PREVENT MANUAL ARMING 

LOCK IN SAFE PRIOR TO FLIGHT 

PREVENT INADVERTENT SAFETY PIN 

REMOVAL 


X 
X 
X 
X 


X 
X 
X 

X 


FIRE 


INITIATE INTERNAL EED 


X 


- 



:^ : > 



Figure 2: Functional Requirements: Conventional Command SADs Vs. Interrupter 

One important aspect of the design criteria 
influenced by the Test Ranges relates to 
manual safing and the Safing Key. The 
SADs must be capable of being manually 
transferred to the Safe condition from the 
Armed state (or any intermediate state), but 
not vice-versa. The Safing Key used for 
manually safing, normally doubles as the 
Safety Pin used to lock the unit in the Safe 
mode, prior to flight. Current range 
requirements dictate that the installed safety 
pin must not be removable if arming is 
inadvertently attempted. 




The Interrupter, shown in Figure 3, has 
been designed to satisfy all Test Range 
requirements relative to firing train 
misalignment, status monitoring, manual 
safing, and the "interlocking" of the Safing 
Key to prevent its removal during 
inadvertent arming attempts. Since the 
device contains no EEDs, the "hand-safe" 
requirement does not apply. 

DESIGN CONSIDERATIONS 

Before undertaking the design of any new 
conventional SAD, certain issues must be 
addressed relative to the functional elements 
of the device. Several of these will be 
defined in the product specification, the 
others will involve trade-off studies to 



Figure 3: 
The Ordnance Transfer Interrupter 

optimize the selection of functional 
approaches. Examples of the specified 
criteria might be as follows: 

a) The general type of SAD (Command or 
Inertial) 

b) Single or dual firing trains? 

c) Hermeticity? 

d) Reversibility of the drive train? 

e) Detonation or deflagration output?. . .etc. 

Choices that can be made by the designer, 
based on such factors as cost and reliability, 
include the type of prime-mover in the 



- 215 - 



drive-train (e.g., solenoids, or springs, or 
..), movable EEDs versus a movable 
barrier, and the type of electrical switching 
components to be used (e.g., rotary or snap- 
action, or ...). 

When generating the design of the 
Interrupter, the ground rules had changed 
slightly, and different choices were to be 
made. This unit was to contain no EEDs, 
therefore, questions regarding firing trains 
only related to the external ordnance transfer 
lines. The issue of SMDC versus CDF lines 
arose, and the propagation characteristics 
were found to be somewhat different for the 
two types. Hermeticity was not specified, 
which simplified the design of the 
interlocking Safing Key, since a welded 
metal bellows "pass-through" was not 
required. It should be noted that the 
Interrupter is environmentally sealed, with 
and without the Safing Key installed. 

Because the unit interrupts fixed transfer 
lines, the interruption must be done by a 
movable barrier. Design choices for the 
barrier included a linear displacement type 
and two rotary displacement types, one a 
disc the other a shaft. Workable designs 
could have been generated based on any of 
these approaches, but the rotating shaft 
option was selected because it was believed 
that it would facilitate the interfacing of the 
barrier with the drive train and the required 
manual safing mechanism. 

A linear solenoid/bellcrank solution was 
chosen for the drive train, rather than a 
more conventional rotary solenoid. The 
selection was based on lower cost, and the 
ability to readily obtain a reversible 90 
degree rotation of the barrier without 
resorting to a gear reduction train. 



INTERRUPTER DESIGN 

As shown in Figure 4, the Interrupter 
housing is a complex rectalinear structure 
with overall dimensions of 5 1/4 L x 3 5/8 
W x 2 7/8 H inches. The lower LH view in 
Figure 4 shows two of the four input/output 
detonator ports (the other two oppose the 
ones shown) in the aft section of the 
housing, and two electrical connectors. One 
of these interconnects with the drive train 
power circuits, the other connector 
interfaces with the remote monitoring 
circuits. The two lower views depict the 
Safing Key, which is located in a cylindrical 
projection on top of the unit. The two 
cylindrical projections at one end of the 
Interrupter house the pneumatic damper 
components. 

Figure 5 shows the drive train components. 
The prime movers are two identical pull 
type solenoids, installed in parallel, one for 
driving the mechanism from the Safe to the 
Armed state, the other for reversing the 
procedure. The plunger in each solenoid is 
pinned to a rod extension on which a roller 
is mounted. A low-inertia rotor is mounted 
on a shaft aligned on an axis normal to the 
solenoid axes and half way between them. 
As shown in Figure 5, the rod-mounted 
rollers contact opposite faces on the Rotor. 
A retraction of the Arm solenoid causes its 
coupled roller to cam-drive the Rotor in the 
clockwise direction (viewed from the rotor 
side of the unit). Conversely, the Safe 
solenoid will drive the Rotor in the CCW 
direction. An overcenter spring, pinned, to 
the Rotor and housing , completes the 
Rotor's full 90 degree rotation after it is 
driven past top-dead-center by either 
solenoid. The overcenter spring is also 
designed to detent the Rotor in either the 
Safe or Armed position. 



- 216 - 







.06 MAX- 
2X .52 UAX- 



^ 



1* 



1-63 

i I 

i | 



3 



5.265 
'5.235" 



J^ 



/ 1 



-©■ 



/ 



-©- 



STATUS VIEWPORT - 



n 



\ 






Figure 4: The Interrupter Configuration and Envelope 



2X .877 
T | 1.412 

. J L 1-382 

-L I 

_] L 



3.637 
3.607 



J i. 



i 
3.902 MAX 

i i 

2.S75 



.1 



J L_^ 











SECTION D-D @7 Y© 

Figure 5: Cross-Sectional Views of the Interrupter 



- 217 - 




Figure 6: The Pneumatic Damper 



An unusual feature of the Interrupter drive 
train is the incorporation of pneumatic 
dampers, which are coupled in-line with 
each of the solenoid actuated pull rods. The 
purpose of these dampers is to prevent the 
rollers from hammering on the rotor, during 
shock and vibration exposure. Each 
damper, as shown in Figure 6, consists of a 
spring-loaded piston riding in the bore of a 
tubular extension on the housing. The 
damper-piston head and rod are sealed by 
dynamic "labyrinth" glands designed to 
minimize friction drag. The damping force 
is controlled by an orifice through the head 
of the piston. Figure 7 shows the calculated 
effect of one of these dampers on the 
rotational velocity of the rotor. The rotor 
shaft, which is mounted in plain bearings at 
each end, is the barrier which blocks (or 
permits) detonation propagation between the 
input (donor) and output (receptor) tips on 
the ordnance transfer lines which are 
coupled to the Interrupter. Two transverse 
apertures in the shaft are aligned with the 
ports to provide propagation paths when the 
rotor is in the Armed state. When Safe, the 
apertures are maintained at 90 degrees to the 
propagation paths. The apertures, which are 
carefully configured slots rather than 
cylindrical bores, are shown in Figure 8. 




\0 20 30 46 SO 60 

Figure 7: 
Damper Effect on Rotor 



(J 88) 




4X FULL 



Figure 8: 
Propagation Apertures in Rotor Shaft 



- 218 - 



The dimensions and configuration of these 
slots evolved during development testing, to 
provide reliable detonation propagation 
across in airgap of almost 1/2 inch. To put 
this simple fact in perspective, the airgaps 
between donor and receptor tips in ordnance 
transfer lines are usually of the order of 40 
to 60 thousandths of an inch. 

As shown in Figure 9 a spur gear is 
mounted on the outboard end of the rotor 
shaft. This gear meshes with a floating 
gear-rack, which is coupled, by means of a 
pin projecting from its back-face, with a 
spring-loaded push rod. Depressing the 
push rod against its spring, by the insertion 
of a Safing key, will thus activate the rack, 
thereby causing the spur gear, and hence the 
rotor shaft, to rotate. This is the mechanism 
for manually driving the interrupter from the 
Arm to the Safe state. A partially slotted 
section in the push rod ensures that it can 
only drive the rack in the Sating direction. 
In other words, the unit cannot be manually 
driven from the Safe to the Arm position. 
The Safing Key that is used with this device 
is a simple bayonet-type pin featuring a 
radial button and integral blade at one end. 
When the key is inserted into the housing, 
its radial button engages a slot which guides 
the blade on the key into engagement with a 
clevis on the end of the push rod. As 
insertion continues, the key depresses the 
push rod, which is also guided by a pin-in- 
slot feature, thereby safing the unit. After 
the key is fully inserted, it is rotated 90 
degrees, where it's button meets a stop. 
From this position, the spring-loaded push 
rod forces the key back, until it's button is 
captured in a detent slot. At this point, with 
the Safing Key detented while holding the 
push rod in a depressed (and rotated) state, 
the Interrupter is locked in the Safe 
condition. 




SATING -PIM 

VJetJei or \>raitd irtU Hsg 
RACK 



Rotor -Co Ha r 



Figure 9: 
Manual Safing Mechanism 

One simple feature on the push-rod permits 
the Interrupter to satisfy a current test range 
requirement which stipulates that, when 
installed, SAD safing keys must not be 
removable when an inadvertent attempt is 
made to arm the device. A notch in the 
push rod is aligned with the pin on the rack 
when the detented Safing Key has rotated 
the engaged push rod through 90 degrees. 
This notch permits a small amount of rack 
movement if an attempt is made to drive the 
rotor from Safe to Arm. The amount of 
rack travel, which corresponds to approxi- 
mately 10 degrees of rotor rotation, is 
sufficient to drive the rack pin into the push- 
rod notch. When trapped in the notch, the 
pin prevents rotation of the push-rod, which, 
in turn, prevents rotation of the Safing Key, 
and hence it's removal from the unit. 

As well as providing the detonation transfer 
barrier and an important part of the manual 
safing mechanism, the rotor shaft serves one 
more purpose, namely that of a flag bearer. 
A two-color disc is mounted on the end of 
the shaft, and viewed through an offset 
window in the housing, for visual status 
monitoring. When the shaft is in the Safe 
position, only the green side of the "Flag" 
disc is observable, when in the Armed 



- 219 - 



position, only the red side is seen. 



DEVELOPMENT 



A pair of passive electrical circuits are 
incorporated in the Interrupter, for remote 
interrogation and monitoring of the unit's 
Safe/ Arm status. These circuits, which are 
shown schematically in Figure 10, are 
alternately closed by sub-miniature snap- 
action electrical switches, actuated by the 
rotor. This design approach was selected 
primarily because of its simplicity, hence 
potential reliability and cost effectiveness, 
compared with the PCB/brush-contact type 
of rotary switches commonly used in SADs. 
The approach was rendered viable by the 
fact that no EED firing circuits require 
switching within the Interrupter. 




ji 

PT02E-8-4P 



J2 
PTD2E-S-3P 



SAFE INPUT 
SAFE RETURN 

ARM INPUT 
ARM RETURN 



SAFE MOMTDR 



MONITOR RETURN 



ARM MONITOR 



Figure 10: 
Interrupter Electrical Schematic 



A comprehensive development test program 
was undertaken, directed towards the design 
characterization and refinement of the 
barrier, relative to its effectiveness, both as 
a block, and as a propagation path in the 
ordnance transfer line. The blocking tests 
were conducted using special fixtures 
capable of precision settings of a range of 
angular misalignments. Short lengths of 
CDF line, with standard detonation end-tips, 
were used to accurately represent the 
transfer lines in these tests. Effective and 
reliable blocking was found with the barrier 
apertures less than 40 degrees out of 
alignment with the output ports. In the Safe 
position, the apertures are misaligned a full 
90 degrees. 

During the transfer tests, several changes 
were made to the barrier aperture 
configuration before reliable propagation 
across the 1 12 inch airgap could be 
achieved. When the optimum aperture size 
and shape appeared to have been derived, it 
was proven by means of a Bruceton series 
of tests. 

A commercially available linear solenoid 
was selected as the drive-train prime mover, 
because of its compact size and advertised 
high pull-in force. During development 
testing, the solenoid proved to be marginal 
in performance at the specified lowest input 
voltage level. This was partly because the 
switch actuator drag forces, on the rotor, 
were higher than expected. Changes were 
made to the switch actuators, and eventually 
the switches themselves were changed, 
resulting in a solution to the problem. In a 
recent design refinement of the Interrupter, 
the solenoids were increased in size to 
substantially enhance the pull-in force 
margin. 



- 220 - 



EXPLOSIYE GAP PROPAGATION 


: Tests per DOD- 


-E-83578 


VIBRATION: 

Frequency 

20 Hz 

20 to 70 Hz 

70 to 800 Hz 

800 to 20000 Hz 

2000 Hz 

Overall 


X and Y Axis 

.026 G 2 /Hz 

4-6 dB per Octave 

0.3 2 GVKz 

-6 dB per Octave 

.051 GVHz 

19.8 GRMS 


Z Axis 
.041 G 2 /HZ 
+6 dB oer Octave 
0.50 G 2 /Hz 
-6 dB per Octave 
.08 G 2 /Hz 
25.0 GRMS 


SHOCK: 






Frecruencv 

100 
100 to 1500 

1500 

3000 


Peak Acceleration 

45 
+5 dB per Octave 

4100 

4100 




BENCH TEST 25 cycle 


test at vacuum ( 


26.8V input) 


TEMPERATURE CYCLING: 8 


cycles -8 5°F and 


+ 180°F 


CYCLE LIFE TEST: 1000 


cycles 




STALL TEST: 32V inDUt 


for 5 minutes and for 1 hour 



Figure 11: The Qualification Test Program 



One area of concern, going into the 
development program, was the possibility of 
the linear drive trains dislodging and 
displacing the detented rotor, when 
subjected to the full range of shock and 
vibration environments specified by the 
customer. No such problems were 
experienced when the units were subjected 
to the dynamic tests, thus indicating the 
effectiveness of the pneumatic dampers. 

QUALIFICATION 

In September of 1993, a group of 
Interrupters successfully completed a 
qualification test program, as defined by the 
customer. Figure 11 shows the tests that 
were conducted, which included temperature 
cycling and a 1,000 cycle life test, as well 
as the dynamic environments. This program 
has qualified the Interrupter for flight at the 
Wallops Island Test Range. 



FUTURE REFINEMENTS 

The Interrupter is a new product, and as 
such, we would be very naive to think that 
it cannot be improved. As already noted, 
the solenoids in the first units were smaller 
than optimum, and the next larger standard 
frame size solenoid is planned for future 
units. 

Already, studies have been made on a cost 
effective installation of a rotary switch, as 
shown in Figure 12. This will reduce 
frictional drag on the rotor, as well as 
provide for the switching of additional 
circuitry. In a current new application for 
the Interrupter, EEDs will replace the input 
transfer lines. The firing circuits for these 
EEDs will be routed from the power input 
connector to the rotary switch, and back 
outside the housing through an additional 
connector. The EEDs would be cabled to 
the additional connector, which will be 



- 221 - 



mounted on the aft face of the housing. 

Another possible design refinement would 
be full integration of the pneumatic dampers 
within the main housing, rather than in the 
tubular extensions. This would have the 
advantage of reducing the overall length of 
the unit, although it might make the 
assembly of the device slightly more 
difficult, therefore, more expensive. 

CONCLUSION 

The Interrupter described in this paper is a 
"patent pending" device which offers a 
significantly lower cost alternative to 
conventional Safe and Arm devices, for 
some applications. The original design was 
limited to applications involving ordnance 
transfer lines interconnecting system 
initiators with independent output devices. 
A recent refinement to the Interrupter, in 
which a rotary switch replaces the original 
microswitches and an additional connector is 
added , permits electrically initiated 
detonators to be installed in the unit's input 
ports. This would allow the Interrupter to 
be used in many more SAD applications. 
The refinements will not eliminate the basic 
advantage that prompted the generation of 
the Interrupter concept in the first place. 
That is, the Interrupter will remain a 
completely passive device, with no internal 
ordnance components, hence it will be 
completely safe to handle. 




ARM SOLENOID 



SAFE MONITOR #1 



L SAFE MONITOR #2 

SAFE CONDITION 




-OET # 1 



SAFE SOLENOID 



ARM MONITOR 



ARMED CONDITION 

Figure 12: 
The Rotary Switch Refinement 



- 222 - 



Sii-22 

6m 



10 



A VERY LOW SHOCK ALTERNATIVE TO CONVENTIONAL. PYROT ECHNICALLY 

OPERATED RELEASE DEVICES 



Mr. Steven P. Robinson 

Senior Mechanical Design Engineer - Research & Technology 

Boeing Defense & Space Group 

Seattle, Washington 



ABSTRACT 

NiTiNOL is best known for its ability to remember a 
preset shape, even after being "plastically" deformed. 
This is accomplished by heating the material to an 
elevated temperature up to 120 degrees C. However, 
NiTiNOL has other material and mechanical 
properties that provide a novel method of structural 
release. This combination of properties allows 
NiTiNOL to be used as a mechanical fuse between 
structural components. When electrical power is 
applied to the NiTiNOL fuse(s), the material is 
annealed reducing the mechanical strength to a small 
fraction of the as-wrought material. The preload then 
fractures the weakened NiTiNOL fuse(s) and releases 
the structure. 

This paper describes the mechanical characteristics of 
the NiTiNOL alloy used in this invention, structural 
separation design concepts using the NiTiNOL 
material, and initial test data. Elimination of the 
safety hazard, high shock levels, and non-reusability 
inherent with pyrotechnic separation devices allows 
NiTiNOL actuated release devices to become a viable 
alternative for aerospace components and systems. 

PSTCOPUCTION 

Explosive bolts and separation nuts have been 
successfully applied for structural release operations 
for over 40 years. These devices were simple, cost 
effective and very reliable. However, the increased 
sophistication, and susceptability, of electrical and 
electronic systems in aircraft, missiles and spacecraft 
has increased the effect of pyrotechnically actuated 
release devices from being a mere nuisance to a 
critical path situation that must be accounted for in 
assuring successful system performance. This has 
elevated the status of structural separation testing, via 



explosive bolts, to a very time consuming and costly 
endeavour. 

Within the last five years, emphasis has been placed 
on finding alternatives to explosive bolts and 
separation nuts. The reason for this change of 
direction is based primarily with the explosive nature 
of these devices. The safety issues, when dealing with 
explosives, add additional costs to assembly, testing 
and storage of aerospace components. The shock 
generated by these devices is becoming a critical 
design consideration because of the sophisticated 
electronics being implemented to lower cost and 
improve system performance. EMI susceptability, 
potential contamination from explosive byproducts 
and limited shelf life are other factors that demonstrate 
explosive structural separation is no longer as cost 
effective and simple to use as in the past 

Historically, non-explosive structural separation 
involved electro-magnetic solenoids or wax actuators 
pulling pins to release the structural elements. These 
are capable of performing the release functions but 
operate at a distinct disadvantage because of the 
slower actuation speed and greater volume and weight 
compared to pyrotechnic devices. 

Since 1986, Boeing Defense & Space Group has been 
actively researching a class of materials known as 
Shape Memory Effect (SME) alloys to provide a 
simple actuation mechanism that will combine the 
best features of both non-pyrotechnic and pyrotechnic 
release technologies. Through this work, Boeing has 
developed proof-of-concept structural release concepts 
based on the shape memory effect characteristic of 
NiTiNOL. These early concepts demonstrated that 
NiTiNOL is capable of achieving most of the design 
goals of eliminating explosives, providing reliable 
performance, and demonstrating multiple operation 
capability. However, these devices were volume 



- 22'5 - 



inefficient and slow compared to existing pyrotechnic 
equivalents. 

To improve the performance of our release design 
concepts, a review of the basic characteristics of 
NiTiNOL was initiated to determine if any properties 
were overlooked that would help reduce the size and/or 
increase speed of operation. This review uncovered the 
fact that "as-wrought" NiTiNOL, prior to annealing, 
is very strong. The ultimate tensile strength can be as 
high as 270 KSL When the NiTiNOL is annealed, 
restoring the crystalline phase structure necessary for 
shape recovery, the ultimate tensile strength is 
reduced by a factor of 2 or more. This fact along with 
other characteristics such as high electrical resistance, 
excellent corrosion and fatigue capabilities led us to 
believe that a simple, effective, and fast NiTiNOL 
mechanical "fuse" separation concept is feasible. 

Using NiTiNOL as a mechanical "fuse" appeared to 
be a simple structural separation concept with few of 
the problems associated with pyrotechnic devices. 

INITIAL CONCEPT DEVELOPMENT 

The first test to demonstrate the NiTiNOL mechanical 
fuse concept was relatively simple. This is shown in 
Figure 1. One end of a NiTiNOL wire was mounted 



NiTiNOL 
wire 



terminal strip 




power 
supply 



weight 



Figure 1 NiTiNOL Structural "Fuse" Test Set-up 

to a terminal strip. This was also the positive 
terminal of a power supply. A weight was suspended 
from the wire. The other end of the NiTiNOL wire 
was tied to the negative terminal of the power supply. 
When power is applied, the NiTiNOL is heated well 
into its annealing temperature zone. The strength of 
the NiTiNOL falls to near zero allowing the dead 
weight to fracture the wire releasing the weight. 



This demonstrated that using NiTiNOL as a 
mechanical fuse as a means of holding and releasing a 
given preload was feasible. However, any structural 
alloy should be capable of accomplishing the same 
task. A comparison chart showing the requirements of 
a mechanical fuse compared to the characteristics of 
NiTiNOL, nichrome, beryllium-copper, and steel is 
given in Figure 2. As shown , NiTiNOL has the best 
combination of properties necessary for a mechanical 
fuse release concept 

The most significant factor is the dramatic change in 
strength capability at elevated temperature. This 
reduction in the tensile strength of NiTiNOL is 
crucial to the preload breaking the structural tie and 
releasing the load. None of the other materials show 
as large a strength reduction at elevated temperature. 

The demonstration of fusing a single NiTiNOL 
element does not automatically demonstrate the idea 
can be scaled up to practical sizes and applications. 
Since conventional separation nuts are capable of 
loads up to 25,000 lbf, the NiTiNOL separation idea 
would also have to be capable of achieving these load 
levels. In order to accomplish this, a relatively large 
number of NiTiNOL fuses would have to be 
incorporated in parallel fashion to increase the load 
carrying capability to levels equivalent to explosive 
bolts and nuts. Multiple NiTiNOL fuse element 
arrangements appear to be the only way to maintain 
large load carrying capability and still have the 
resistance of the elements high enough for efficient 
electrical heating 

However, the large number of elements, if they were 
all heated at the same time, would require a 
prohibitive amount of electrical power. This is not 
possible with existing power system ratings on 
today's aerospace systems. A NiTiNOL fusible 
element requires a low voltage, high current electrical 
pulse to efficiently heat the element in the shortest 
amount of time. 

A review of separation time requirements showed a 
large percentage of release operations do not require 
separation times less than 10 milliseconds as is 
typical of pyrotechnically operated separation nuts and 
bolts. The near instantaneous release time is just a 
consequence of utilizing explosives in the separation 
device. This fact allows us to reduce the number of 
elements being heated at any one time to a minimum 
because separation time is not always critical. 

By applying this fact to the NiTiNOL fuse concept, 
we can reduce the instantaneous power requirement to 
manageable levels. This is shown in figure 3. The 5 
element group is mechanically attached in parallel to 



- 224 - 



distribute the load and increase the overall load 
carrying capability. The element lengths are all 

Material Property Comparison Chart 



Properties 


fuse 
ream'ts 


NiTiNOL 


Steel 


304 
Steel 


Beryllium Nichrome 

Cooper Ni70.Ci2S.Mher! 


electrical 

resistivity 

(microhm-cm) 


High 


100 


100 


72 


20 


134 


tensile strength 

(KSI) 
(room temp.) 


High 


250 


210 


110 


115 


150 


tensile strength 

(KSI) 
(1000 deg F) 


Low 


>25 


150 


50 


85 


100 


corrosion 
resistance 


High 


High 


High 


High 


Med. 


High 


fatigue 
resistance 


High 


High 


High 


Med. 


High 


- 


thermal 
conductivity 
(BTU/hr-ft-F) 


Low 


10.4 


10.4 


« 


120 


,4 



Comparison of NiTiNOL to other structural alloys 
Figure 2 



stationary structure 


NiTiNOL fusible 




f MYli ir 


\ y element (5 pi) 


Rl 
B 


* 1 
R3 

1 rs 


R 

■71 „ 


.5 


Jp 


Pa 


ft 


^1 power 
1 supply 


released 
structure 






R1>R2>R 


3>R4>R5 



NiTiNOL Element Sequential Separation Concept 

Figure 3 

different to produce a uniformly increasing resistance 
value range. The elements are wired in parallel. When 
current flows through the elements, the shortest 
NiTiNOL fuse draws the most current, heats the 
fastest and fractures first. Now the load is carried 
among fewer elements. This increases the stress 
levels in each element. The power is also shared 
among fewer elements causing the elements to heat 
even faster. This cascading effect fractures each higher 
resistance element until the last one in the group 
separates. Figure 4 shows an idealized trace of the 
cascading separation effect. The increasing resistance 
of each successive element causes a distinctive 
zipping effect 



To further increase the load carrying capability, 
multiple groups of these subsets of NiTiNOL fusible 
elements can be arranged to be released in series. As 
soon as the last element in the first group separates, 
power is transfered to the next group of elements, 
thereby continuing the separation sequence. 




NiTiNOL Fusible Element Release Sequence 

Figure 4 

The number of element groups can be increased to 
accomodate a wide range of loading conditions. The 
time-to-separate requirement must be addressed to 
assure there is no impact to the overall separation 
operation. 

However, the increase in the separation time may not 
be critical if a two step separation approach is taken. 
An "arming" operation could take place which would 
release the majority of NiTiNOL fuse elements. This 
would leave a minimum number of elements to 
maintain the structural attachment. When actual 
separation occurs, the power required and separation 
time will be kept to a minimum due to the minimum 
number of elements left to fuse open. This concept 
allows a large number of structural elements to be 
maintained across the joint satisfying a wide range of 
available power, time-to-separate, and structural load 
cases. 

This concept is postulated for large separation joints 
such as payload fairings and other large linear 
structural interfaces. In fact, the majority of this work 
was performed in anticipation of the next generation 
heavy lift launch vehicles (HLLVs). 

As a result of this initial work, a patent (#5,046,426) 
has been awarded to The Boeing Company. 

NASA/JSC SEQUENTIAL SEPARATION TEST 

Using the concept described in the previous section, 
Boeing Defense & Space Group was contracted by 



- 225 - 



NASA/JSC to perform a feasibility experiment 
demonstrating that a NiTiNOL sequential structural 
separation system is capable of loads in the range 
needed for commercial applications. Since this was a 
small experiment, a candidate separation load was 
assumed to be 5000 lbf. This would provide a 
reasonable loading condition without imposing extra 
costs. 

The basic design concept is shown in figure 5. To 
expedite the experimental hardware fabrication, we 
utilized NiTiNOL strip, 1.4" w x 0.004" t, that was 
available in-house, as part of our ongoing IR&D 
effort. Although the dimensions of the strip was not 
optimized for this experiment, we felt valuable 
information on laser cutting of NiTiNOL and 
operation of this patented NiTiNOL non-pyrotechnic 
release concept could be achieved. 

The structural members were fabricated from 4.0" dia. 
molybdenum disulfide impregnated nylon. This 
provided an inexpensive, electrically isolating 
material capable of handling the 5000 lbf projected 
load. Mounting studs were attached to the center of 
the nylon parts to provide sufficient grip length for 
installation onto an Instron tensile test machine. The 
NiTiNOL fusible element strip was installed across 
the interface between nylon members. The NiTiNOL 
fusible element member was attached by two(2) rows 
of 32 each 6-32 fasteners. These were installed into 
tapped holes in the nylon parts. The load across each 
fastener was 125 lbf max. The one concern was 
whether the attachment holes, in the NiTiNOL strip, 
were strong enough to react the tensile load without 
tearing out. 

The cutout pattern and slots, defined the five (5) 
NiTiNOL fusing elements per each of the eight (8) 
groups. The cutouts were produced by a high powered 
laser cutting system located in the Boeing Materials 
Technology Laboratory located in Renton, Wa. 
Utilizing computer controlled laser cutting provided 
several benefits. Unique patterns can be cut into the 
strip with great accuracy. This also provides a high 
degree of dimesional repeatability, critical for some 
operations. Laser cutting also provides a way to 
minimize the area of the heat affected zone which 
would compromise the large differential strength 
characteristic of NiTiNOL from its unnannealed state 
to its annealed state. 

The structural separation operation uses an electrical 
circuit that applies battery power to opposite pairs of 
fusing elements. This assures a symmetrical release 
of the load minimizing any off-axis unloading 
situations resulting in excessive tip-off rates. As the 
last elements of the first two groups are fused opened, 
battery power is switched to the next pair of fusing 



element groups. This continues until all fusing pairs 
of element groups have been severed. The circuit 
diagram is shown in figure 6. To expedite the circuit 
design, automobile starter solenoids were utilized in 
the circuit design to transfer battery power between 
NiTiNOL fusing element groups. 

When switch SI is engaged, 28 V is applied to the 
first starter solenoid closing the circuit and applying 
12 V battery power across opposite groups of 
NiTiNOL elements. The 4 ohm resistor prevents the 
second solenoid from engaging until the last 
NiTiNOL element, from the first 2 groups, has fused 
open. Battery power is then switched to engage the 
second solenoid, which in turn applies battery power 
to the next pair of opposite NiTiNOL element 
groups. This continues until the last NiTiNOL 
elements are fractured releasing the structural load. 

EXPERIMENTAL TEST RESULTS 

Using 0.004" thick foil for this test generated concern 
that the foil might fail in the attachment holes at the 
required load of 5000 lbf. A load test was performed 
to determine the maximum load capability. As 
predicted, the failure occured in the mounting holes at 
approximately 3200 lbf. It was obvious that a goal of 
5000 lb was not possible with the current material. 
However, no alternative was available to support 
testing. Therefore, it was recommended that the test 
load be reduced to 2000 lbf. This would still 
demonstrate the feasibility of this technology with 
the current experimental hardware at a realistic load 
value. 

One test was performed to demonstrate feasibility. 
The test article was mounted on an Instron tensile test 
machine with full scale readout of 5000 lbf. The load 
was uniformly increased to 2000 lb indicated. As the 
load reached the test level, switch SI was closed 
applying power to the first solenoid. The first pair of 
NiTiNOL element groups fused opened in 156 
milliseconds, the second pair in 182 msec, the third 
pair in 164 msec, and the last pair in 194 msec. The 
total time to release was 0.838 seconds. The trace of 
the release operation is shown in figure 7. As the 
oscilloscope trace shows there was some bounce of 
the solenoid contacts generating some delay of power 
to the NiTiNOL elements, increasing the apparent 
separation time. 

The circuit performed as designed. Once the switch 
was engaged, the application of battery power was 
autonomous and continuous. This resulted in a very 
simple circuit capable of transferring high current 
pulses as many times as needed. 



- 226 - 



NITINOL RELEASE TEST F I X TURE ( FULL Y ASSEMBLED 



NITINOL FUSIBLE ELEMENT STRIP 



eeee - 



«rH$^ 



1HREAOED STUO 




STRUCTURE CNYLON ) 



8 9 e e HG/fl* 

$ $ $ f 4H 



NITINOL FUSIBLE 
ELEMENT GROUP (I QF 8) 



STRUCTURAL ATTACHMENT 
HOLES 




Figure 5 NASA/JSC Sequential Structural Separation Demonstration Experiment 



released structure 



)))))))) 



starter 1 
solenoid) 
(1 of 4^ 




stationary 
structure 

4 ohms 

(typ) 



+ 
12 V battery 



asi 



Til ■ 



si 



s~ 



power supply 



Figure 6 NASA/JSC NiTiNOL Fusing Element Electrical Circuit 



SUMMARY 

Boeing Defense & Space Group believes this 
technology could provide a viable alternative to 
explosive separation systems utilizing linear shaped 
charges to weaken and fracture a structural joint, such 



as those on large payload shrouds. Further research 
into the possibility of gradually releasing the preload, 
prior to full separation, offers design possibilities that 
could reduce the shock of separation, power usage, 
and separation time even further. 



- 227 - 



Although this type of release concept may require a 
unique electrical system, such as dedicated on-board 
batteries, the changes appear to be minimal and 
simple to implement. 




Time-to-Release Separation Test - Oscilloscope 
Traces (2000 lbf preload) 

Figure 7 



If an existing electrical system, capable of operating 
pyrotechnic devices with 5A DC max. current output 
is the only source of power, work was accomplished, 
under contract with the Naval Research Laboratory, to 
develop such a release device, for spacecraft use, based 
on this invention. This work is described in the 
following section. 

NiTiNOL FUSIBLE LINK RELEASE DEVICE 
(NAVAL RESEARCH LABORATORY) 

The Naval Research Laboratory contracted with 
Boeing to develop a NiTiNOL based mechanism to be 
included as part of the Advanced Release Technologies 
(ARTs) program. The requirement of being able to 
interface with an existing 28V/5A spacecraft power 
bus system needed a different design approach than the 
NASA/JSC concept. In order to accompliash 
separation of a 2000 lbf preload within 0.250 second 
using a limited power budget, we used a single 
NiTiNOL fusible element, in conjunction with a 
large mechanical advantage, as the active member to 
accomodate a 2000 lbf preload. The basic concept is 
shown in figure 8. 

The overall size of the device is 3.50" x 3.50" x 1.5". 
Although larger than conventional separation nut 
designs, the size envelope is small enough to be 
useful in many separation operations. Future design 
iterations can conceivably reduce the size even further. 

The most significant change between the NASA/JSC 
concept and this concept is the use of a 9:1 step-down 
transformer. The transformer, along with the DC/AC 
converter electronics, allows the device to operate 
with an existing 28V/5A max electrical power bus 
system. This system is typical of current spacecraft 
designs. The electronics converts 28V DC to 28V AC 
at 100 KHz. The step-down transformer converts the 
chopped 28 V/5A AC to approx. 3.1 V/45 A AC 
power. The high frequency of the chopper electronics 
allows us to use the smallest transformer possible. 
The total power usage has not changed. However, it 
has been converted to a more useable form for 
efficient heating of the NiTiNOL fusible element. 

The design concept provides a mechanical advantage 
of approximately 24:1. This enables a NiTiNOL 
fusible link, sized for 150 lbf, to be able to withstand 
a 2000 lbf preload. In fact, the fusible link is sized for 
3600 lbf. This corresponds to a positive margin of 
safety of approximately +1.75. The NiTiNOL fusible 
link design is also shown in figure 8. In order to 
minimize the transformer lead lengths, the design of 
the fusible link is a U-shape configuration allowing 
both transformer leads to be on the same side of the 
release device. This also provided the added benefit of 



- 228 - 



STOWE 



RELEASED 



TENSION LINK 



HOUSING 



TORSION SPRING 
(1 OF 2) 




Figure 8 



TRANSFORHER- 

NiTiNOL Fusible Link Release Device (NRL) 



doubling the strength of the NiTiNOL mechanical 
fuse without increasing the overall size of the release 

device. 

The release device configuration is straightforward. 
Two (2) spring loaded jaws are closed to capture the 
tension link. The NiTiNOL fusible link is installed 
on two phenolic blocks at the ends of the jaws. The 
jaws have a step at the bottom where the tension link 
engages the jaws. When the preload is applied 
through the tension link, the step creates a 0.10" 
moment arm. The NiTiNOL fusible link has a 
moment arm of 2.4 ". This creates a 24:1 mechanical 
advantage. The allows a relatively small fusible link 
to be employed against a substantial preload. Figure 9 
describes the geometry in greater detail. 



In order to keep the device weight and volume to a 
minimum, the transformer, designed to operate at a 
frequency of 100 KHz, was used. The transformer size 



NiTiNOL Fusible Link Reaction Force Geometry 



tension link 



20q01b NiTiNOL 

fusible link 




2.40" 



lb preload 
applied 

0.10" 



pivot pt 



(2000x0.10")= 2.40" x F 



F = 851b 



Figure 9 



- 229 - 



was 1.3" L x 1.0" W x 0.25" t. Mounting the 
transformer to one of the jaws kept wire lengths to a 
minimum. This prevented inductance from becoming 
a problem. Too much inductance would reduce the 
amount of power through the NiTiNOL fusible link 
compromising the heating of the NiTiNOL and the 
performance of the device. 

The preload is applied by threading a bolt into the top 
of the tension link assembly. As the bolt was 
torqued, the tension link would be pulled up and 
engage the jaws. The moment generated by the 
tension link against the jaw tries to force the jaws 
apart. The NiTiNOL fusible element reacts this torque 
until the NiTiNOL is electrically heated. When this 
occurs, the link becomes structurally weak and 
fractures, allowing the jaws to spring out and the 
tension link to be extracted. 

The DC to AC chopping circuit is on a separate board 
and can be installed in any convenient place. It can 
also be installed on the release device housing itself. 

TEST RESULTS 

The release device was proofloaded to 2000 lbf 
without separation. When power was applied, the 
release device demonstrated release time less than 200 
msec. Several tests were also conducted at lower 
preloads. The effect of the preload variation on release 
time was apparent. It showed that lower preload 
values yielded higher release times. At no preload, the 
release time was approximately 50% greater. This 
required the NiTiNOL to be heated to near the melting 
temperature. Under actual conditions, this zero preload 
situation would be very remote. 

A longterm loading effects test was also performed on 
a NiTiNOL fusible element. This was to determine if 
any stress relaxation or creep phenomenon was 
present using NiTiNOL. The link was mounted in a 
fixture with a simulated preload. This was stored for 
approximately six (6) months. Measurements were 
taken on a daily basis. No significant increase in 
length was observed for the entire 6 month period. 

Separation tests confirmed the ability of a NiTiNOL 
fusible link release device to maintain and release a 
2000 lbf preload reliably. Testing at NRL is ongoing. 
Initial testing shows separation times are consistently 
within 50 msec. However, this is dependent on the 
same power and preload being applied during each 
separation test. 

SUMMARY 

This non-pyrotechnic release concept demonstrated 
that a single NiTiNOL fusible element can reliably 



hold and release a given preload using a typical 28V, 
5A electrical bus system. Even with the apparent 
dependency of release time to preload, this can be 
attributed to the limited power available. The effect 
can be minimized by proper sizing of the NiTiNOL 
fusible link and optimizing the heating to the power 
availability. In addition, the shock of separation was 
insignificant. There is no contamination or safety 
issues associated with this device. 

The release device is completely reusable except for 
the NiTiNOL fusible link. This feature allows the 
same device to be operated many times during ground 
testing, and still be available as the flight unit. The 
benefits of this device are shown in figure 10. 



Benefits 



1) Non-pyrotechnic 

2) Fly-as-tested capability 

3) Little or no separation shock 

4) No shelf life limitations 

5) No safety hazards 

6) No EMI susceptability 

7) Fast separation time 

8) No contamination potential 



Figure 10 



NiTiNOL Beneffits Chart 



These features can provide a very cost effective 
product especially if extensive ground testing is 
contemplated. The cost of the NiTiNOL material does 
not appear to be a limiting factor because commercial 
usage continues to increase as more applications are 
realized. As usage increases, the material price will 
decline accordingly. 

CONCLUSION 

Load capability and separation times demonstrated by 
these concepts show that NiTiNOL fusible element 
based devices, using this Boeing patent, have the 
potential to achieve the same performance as 
pyrotechnic devices. This can be accomplished 
without the detrimental effects attributed to the use of 
explosives, 

Boeing Defense & Space Group feels this technology 
will provide a much needed reduction in safety related 
and shock environment issues involving aerospace 
vehicles. Reducing shock environmental requirements 
imposed on vehicle sub-systems and components will 
play a major role in reducing vehicle development 
costs. The costs associated with handling, storage and 



- 230 - 



assembly of pyrotechnic devices can be practically 
eliminated if this technology can be developed to its 
fullest capability. 

Both of the concepts, described previously, offer both 
ends of the design spectrum that is possibile using 
this simple technology. Many design alternatives can 
be created if the drawbacks, associated with 
pyrotechnic devices, can be eliminated. We understand 
this and are continuing to improve the basic concepts 
described here. 

One of the most intriguing design possibilities is the 
two-step arming/separation function described 
previously. This idea offers unique advantages and 



design flexibility that provides the designer with 
options not possible with conventional pyrotechnic 
systems. This ability to slowly release large preloads 
all but eliminates the heavy shock environment 
imposed on the surrounding structure. This can be 
accomplished without jeopardizing the actual release 
function. 

The future of non-pyrotechnic structural separation, 
based on this patent, will be expanding. The 
capabilities offer so many advantages that this 
technology will become a major part of structural 
separation for the next generation of aerospace 
vehicles. 



- 231 - 



INVESTIGATION OF FAILURE TO SEPARATE AN INCONEL 718 FRANGIBLE NUT 



<, lit- h 

70O0 



William C. Hoffman, III 

Carl Hohmann 

NASA Lyndon B. Johnson Space Center, Houston, TX 



Abstract 

The 2.5-inch frangible nut is used in two places to 
attach the Space Shuttle Orbiter to the External Tank. It 
must be capable of sustaining structural loads and must 
also separate into two pieces upon command. Structural 
load capability is verified by proof loading each flight nut, 
while ability to separate is verified on a sample of a 
production lot. Production lots of frangible nuts 
beginning in 1987 experienced an inability to reliably 
separate using one of two redundant explosive boosters. 
The problems were identified in lot acceptance tests, and 
the cause of failure has been attributed to differences in 
the response of the Inconel 718. Subsequent tests 
performed on the frangible nuts resulted in design 
modifications to the nuts along with redesign of the 
explosive booster to reliably separate the frangible nut. 
The problem history along with the design modifications 
to both the explosive booster and frangible nut are 
discussed in this paper. Implications of this failure 
experience impact any pyrotechnic separation system 
involving fracture of materials with respect to design 
margin control and lot acceptance testing. 

Introduction 

The 2.5-inch frangible nut is used in the Space 
Shuttle Program to attach the Orbiter to the External Tank 
at two aft attach points as shown in figure 1. Structural 
loads illustrated in table 1 are carried by each frangible 
nut. Upon completion of Space Shuttle Main Engine 
cutoff, at approximately 8 minutes, 31 seconds after 
Shuttle launch, the Orbiter is separated from the External 
Tank by initiation of pyrotechnics at the forward and aft 
attach points. Aft structural separation is accomplished 
by fracturing each of four webs on the two frangible nuts, 
as illustrated on figure 2. Separation is accomplished by 
initiating one or both of the -401 configuration booster 
cartridges shown in figure 3. The Orbiter frangible nuts 
are safety critical devices which are required to reliably 
operate for Shuttle crew safety. Production lots beginning 
in 1987 experienced an inability to operate reliably with 
the performance margins demonstrated in the original 
qualification. An intensive failure investigation followed 
which has identified the Inconel 718 used in the frangible 



nuts as the cause of the failures. The manufacturer of 
Inconel 718 forgings used in the qualification and initial 
lots of frangible nuts for the Shuttle Program went out of 
business, and NASA was forced to solicit new sources for 
the frangible nuts. The change in Inconel 718 suppliers 
and the differences in the characteristics of the material 
led to a performance degradation. 

The first section of this paper discusses the original 
qualification program and the original Inconel 718 
material and chemical properties. The second section of 
the paper discusses the failure analysis performed by 
NASA and the resultant design solution arrived at through 
iterative testing. 

Design and Qualification History of 
2t?-inch Frangifrte Nut 

The 2.5-inch frangible nut is designed with two 
primary requirements. The first requirement is that the 
nut have the capability to carry structural loads with 
specified margins against material yield and rupture. The 
second requirement is that the frangible nut reliably 
separate into two pieces when either one or both booster 
cartridges are initiated. Inconel 718 was selected for the 
frangible nut due to the combined high material strengths, 
and to its resistance to creep and corrosion. The 
qualification matrix shown in table 2 illustrates the type 
and number of tests performed to demonstrate reliable 
operation in the presence of flight and ground 
environmental conditions. The performance margin was 
demonstrated using nominal booster cartridges in 
frangible nuts whose web thicknesses were increased 
above the maximum allowable by 20% as shown on the 
-101 margin nut in figure 4. Shuttle Program 
requirements dictate a margin demonstration of 15%, but 
the additional 5% margin was chosen to gain confidence 
in the frangible nut design. All margin tests were 
successful. Design, development, and test of the 2.5-inch 
frangible nut were conducted under NASA contract 
NAS 9-14000 and results of the qualification were 
reported in document CAR 01-45-1 14-0018-0007B 1 . 

Material Configuration of the Original Manufacturer's 
2.5-inch Frangible Nut 

The supplier of the qualification nuts and boosters 
procured Inconel 718 which was manufactured to meet 
AMS 5662 2 . A compilation of chemical data, material 
properties, and typical microstructure grain size for a 
representative Inconel 718 heat lot used in the original 



- 233 - 



PAGL 



E £it, 



^^;ViXNALLY BL 



- A ■'■?'/ 



manufacturer's frangible nuts is shown in table 3. No 
additional restrictions were placed on the Inconel 718 
other than requiring compliance with AMS 5662 . 
Figure 5 is a representative micrograph of the original 
manufacturer's Inconel 718 shown at a magnification of 
100X. 

Based upon the successful qualification program, the 
design was considered complete and production contracts 
were issued to support Shuttle flights. 

Frangible Nut Production Failures 

NASA solicited new manufacturers of the 2.5-inch 
frangible nut in 1987 in order to develop additional 
sources of supply for the Shuttle Program. Two 
qualification contracts were issued with the intent of 
demonstrating the new manufacturer's processes. The 
second manufacturer was awarded NASA contract 
NAS 9-17496 and the third manufacturer was awarded 
NASA contract NAS 9-17674. During qualification 
testing performed under NAS 9-17496 3 , in accordance 
with table 4, failures were encountered during frangible 
nut margin tests. The frangible nut failed to sever the 
outer web, web number 4 as shown in figure 6, when fired 
using a single booster cartridge. Further testing resulted 
in a successful separation using a margin nut with webs 
15% over the maximum allowable thickness. In an effort 
to establish performance margin for the frangible nuts and 
booster cartridges, the weight of RDX in the booster 
cartridges used in margin tests was reduced by 15%, and 
nominal frangible nuts were used instead of nuts with 
120% webs. Three margin tests successfully separated 
using 85% charge weight booster cartridges and nominal 
frangible nuts as shown in table 4. 

Material properties, chemical data, and micro- 
structure grain size for the Inconel 718 are shown in table 
3. The Inconel 718 heat lot number for the NAS 9-17496 
qualification lot is 9-11446. Figure 7 illustrates a 100 X 
micrograph taken for heat lot 9-11446. There is a 
dramatic difference in the precipitate distribution for heat 
lot 9-11446 as compared with the original manufacturer's 
Inconel 718 micrograph shown in figure 5. 

The second manufacturer was authorized to produce 
additional frangible nuts based upon successful 
completion of the qualification program. The second lot 
of frangible nuts, Inconel 718 heat lot 9-10298, 
experienced an inability to separate under zero preload 
using a single booster cartridge. The gap developed from 
the single booster cartridge firing, illustrated in figure 6, 
was measured to be less than 0.100" for the failed unit. 
Web numbers 1 and 2 were fractured while web numbers 
3 and 4 did not experience any cracking. Table 5 shows 
the chronology of tests performed to understand the 
failure cause and develop a means of overcoming the 
problem. A design solution was arrived at through the 
test series which consisted of modifications to both the 
frangible nut and booster cartridge. 

NASA's first response to the failure was to redesign 
the booster cartridge to provide additional charge to 
overcome the resistance to separate. Booster cartridge 
internal cross sectional area was increased in increments 



of 5% until successful separation was achieved. In the 
course of performing the above tests, the nut was 
observed to "clamshell" open until the outer ledge gap, 
shown in figure 6, was reduced to 0.00". The frangible 
nut outer ledge was machined to provide additional 
rotational motion for web number 4 (the outer web) and 
the modification to the frangible nut is illustrated in figure 
8. The modified frangible nut was identified as a -302 
configuration. An additional change was made by 
loading the nominal charge weight into the bore of a 
booster cartridge body which had been increased in cross 
sectional area by 20%. An example of this booster 
cartridge is shown by the -402 configuration in figure 3. 
By combining the two modifications, the frangible nut, 
which was unable to separate under zero preload using a 
single booster cartridge with 1950 mg of RDX, 
successfully separated with no increase in the explosive 
weight of RDX or no reduction in the web thickness 4 . 

Material properties, chemical data, and 
microstructure grain size for Inconel 718 heat lot 9-10298 
are shown in table 3. A 100X micrograph for heat lot 
9-10298 is shown in figure 9. Heat lot 9-10298 is 
markedly more resistant to separation than the 
qualification heat lot 9-11446. 

The third manufacturer of 2.5-inch frangible nuts 
operating under NAS 9-17674 used Inconel 718 from heat 
lot 9-11446 in its qualification test program. Heat lot 
9-11446 is common to the heat lot used in the 
qualification program performed by the second 
manufacturer under NAS 9-17496. The third 
manufacturer began qualification testing in accordance 
with the test matrix shown in table 6. Testing began with 
a margin nut which, at NASA's request, had a web 
thicknesses 20% over the maximum allowable thickness. 
The 120% margin nut is represented in figure 4 by the 
-101 configuration. The 120% margin nut failed to 
separate. The margin test was selected due to experience 
with failures in margin tests during the second 
manufacturer's qualification test program. 

The failure to separate the frangible nuts using single 
booster cartridges under zero preload or to demonstrate 
margin using frangible nuts with overthick webs raised 
concern at NASA over the new manufacturers* booster 
cartridge performance. Potential causes in degradation of 
the RDX detonation output were investigated by chemical 
and physical analysis of each lot of RDX used by each 
manufacturer. No evidence of degradation was found. 
NASA then initiated a test program to inves'igate whether 
the original manufacturer's booster cartridges performed 
differently from new production lots. The first test 
consisted of firing a frangible nut from the original 
manufacturer under zero preload conditions using a 
booster cartridge from recent production. The second test 
involved firing a frangible nut from the second 
manufacturer under zero preload conditions using a 
booster cartridge from the original supplier. The original 
supplier's frangible nut separated using a new 
manufacturer's booster cartridge, and the new 
manufacturer's frangible nut did not separate using the 
original supplier's booster cartridge. These tests indicated 

- 234 - 



that the booster cartridge was not the cause of the 
frangible nut failure to separate. 

Further qualification testing under NAS 9-17674, 
illustrated in table 6, resulted in failures to separate under 
zero preload conditions even though three frangible nut 
margin tests were conducted under preload conditions 
using booster cartridges loaded with 85% of the nominal 
charge weight. The failure of the zero preload, single 
booster cartridge frangible nut test resulted in the 
frangible nut opening until the outer ledges contacted and 
the outer ledge gap, illustrated in figure 6, was reduced to 
0.00". All of the third manufacturer's nuts were modified 
to remove the outer ledges, illustrated in figure 8, thus 
providing more rotational freedom for the outer web 
during a single booster cartridge firing. The third 
manufacturer resumed the sequence of tests described in 
table 6 without failure following the frangible nut 
modification 5 . The modifications to the frangible nut 
were a result of tests performed during the failure 
investigation matrix shown in table 5. Material 
properties, chemical data, and microstrucure grain size 
data for the Inconel 718 used in the third manufacturer's 
qualification lot are shown in table 3, and the 100X 
micrograph of the material heat lot is shown in figure 10. 

Discussion 

In each of the above qualification and production 
heat lots, the Inconel 718 was produced in accordance 
with AMS 5662. The material properties, yield strength, 
tensile strength, elongation and reduction in area are 
illustrated in table 3. Although a significant difference is 
exhibited in Charpy impact strength 6 between recent 
production lots and the original Inconel 718, reference 
table 3, no correlation between Charpy impact strength 
and frangible nut performance has been made. A NASA 
test 7 using material having impact strength of 15 and 
ultimate tensile strength 191.1 ksi, 0.2% offset yield 
strength of 168.2 ksi, elongation of 16.0%, and reduction 
of area of 27.0% resulted in failure when fired using a 
single booster cartridge and under zero preload. The 
exact combination of chemical, microstructural, and 
physical data required to assure successful separation of a 
heat lot of Inconel 718 under zero preload conditions 
using a single booster cartridge has not been defined. 
Additional test programs are underway at NASA to 
further understand the cause of failures for the frangible 
nuts produced under NAS 9-17496 and NAS 9-17674 and 
to define what characteristics in the Inconel 718 are 
critical for successful operation of the frangible nuts. The 
investigations focus on material property variations in 
Inconel 718 and on efficiency of coupling explosive 
potential energy into the fracture of the four webs. 

Future production of frangible nuts will be assessed 
using additional destructive lot acceptance tests to assure 
reliable operation of the flight hardware. At this time, no 
quantitative test exists to differentiate Inconel 718 as 
acceptable or unacceptable for use in flight nuts short of a 
full scale destructive performance test. If failure occurs at 
that point in production, the products are in final delivery 
status and no rework is possible. 



Conclusions 

The most significant conclusions from the failure 
investigations which NASA has performed on the 2.5- 
inch frangible nuts are as follows: 

A. Specification of Inconel 718 per AMS 5662 is not 
adequate to guarantee successful separation of the 
frangible nut using the original design booster cartridge. 

B. No single chemical or material property currently 
measured is an adequate gage of whether the Inconel 718 
used in a frangible nut will result in failure or success 
during perfomance tests. 

C. The critical nature of the 2.5-inch frangible nut 
mandates extensive testing be performed on each 
production lot to demonstrate operational response and 
performance margin. 

References 

1. Contract NAS 9-14000, Document Number CAR 
01-45-114-0018-0007B, "Qualification Test Report for 
Frangible Nut, 2-1/2 Inch and Booster Cartridge," 
Released May 29, 1980. 

2. AMS 5662 Revision F, "Alloy Bars, Forgings, and 
Rings, Corrosion and Heat Resistant, 52.2Ni - 19 
Cr - 3.0Mo - 5.1(Cb + Ta) - 0.90Ti - 0.50A1 - 18Fe, 
Consumable Electrode or Vacuum Induction Melted 
1775°F (968°C) Solution Heat Treated," Issued 
September 1, 1965, Revised January 1, 1989. 

3. Contract NAS 9-17496, Document Number 3936-10- 
301-401, Revision A, " Qualification Test Report for 
2.5-inch Frangible Nut, NASA PN SKD26100099-301 
and Booster Cartridge, NASA PN SKD26100099-401," 
January 15, 1991. 

4. Contract NAS 9-17496, Document Number 
RA-468T-B, "Acceptance Data Package for 2.5-inch 
Frangible Nut, NASA PN SKD26100099-302," 
June 5, 1992. 

5. Contract NAS 9-17674, Document Number 
0718(03)QTR, "Qualification Test Report 2.5-inch 
Frangible Nut with Booster Cartridge, Used on the 
Space Shuttle Aft Separation System," June 19, 1992. 

6. American Society for Testing and Materials Standard 
E23-88, "Standard Methods for Notched Bar Impact 
Testing of Metallic Materials, Type A Specimen." 

7. NASA Test Report 2P333, "Frangible Nut Test 
Program." 

Acknowledgements 
The authors wish to acknowledge the work of 
Ms. Julie Henkener, materials engineer for Lockheed 
Engineering and Sciences Company, Houston, Texas, in 
preparing, reviewing, and interpreting the Inconel 718 
metallurgical data throughout the frangible nut failure 
investigation. 



- 235 - 



TABLE 1 
2.5 INCH FRANGIBLE NUT STRUCTURAL LOAD REQUIREMENTS 





Limit Load 


Ultimate Load 


Proof Load 


Axial Load 


415,270 


581,400 


456,800 


(Lbs) 








Moment 


53,275 


75,215 


59,097 


(in-Lbs) 









TABLE 2 

ORIGINAL MANUFACTURER QUALIFICATION TEST MATRIX FOR 

2.5 INCH FRANGIBLE NUT AND BOOSTER CARTRIDGE 



Test 



Nut 

Group 

NO 



Functional 

Temp 

<-F) 



Preload Booster 
Tension Mount 
(lbs) (in-lb) 



Cartridges 
Dual/Single 



Functional 
(pass/fail) 



Room Temp. 


A 


70-F 


240,000 





Single 


Passed 


Firing 


A 


70-F 


240,000 





Single 


Passed 




A 


70-F 


240,000 





Single 


Passed 




A 


70-F 


240,000 





Dual 


Passed 


High Temp. 


B 


200-F 


240,000 





Single 


Passed 


Firing 


B 


200-F 


240,000 





Single 


Passed 




B 


200-F 


240,000 





Single 


Passed 




B 


200-F 


240,000 





Dual 


Passed 


Low Temp. 


C 


-65-F 


240,000 





Single 


Passed 


Firing 


C 


-65-F 


240,000 





Single 


Passed 




C 


-65-F 


240,000 





Single 


Passed 




C 


-65-F 


240,000 





Dual 


Passed 



Low Temp. 
Firing with 
Limit Axial Load 



Room Temp. 
Firing with 
Zero Preload 

Margin Demo. 
Firing 



E 
E 
E 
E 



-65-F 
-65-F 
-65-F 
-65-F 



378,000 
378,000 
378,000 
378,000 



65,200 
65,200 
65,200 
65,200 



Single 
Single 
Single 
Dual 



F 


70-F 








Single 


F 


70-F 








Single 


F 


70-F 








Single 



G* 
G* 
G* 



70-F 
70-F 
70-F 



240,000 
240,000 
240,000 









Single 
Single 
Single 



Passed 
Passed 
Passed 
Passed 

Passed 
Passed 
Passed 

Passed 
Passed 
Passed 



Group G nuts had web thicknesses 120% the nominal maximum allowable 



- 236 - 



TABLE 3 
MATERIAL PROPERTIES, CHEMICAL DATA, AND MICROSTRUCTURE GRAIN SIZE FOR INCONEL 718 USED 

IN FRANGIBLE NUTS BY MANUFACTURERS 





ORIGINAL 


NAS9-17496 


NAS9-17674 


NAS9-17496 




MANUFACTURER 


QUALIFICATION 


QUALIFICATION 


PRODUCTION 




HEAT LOT 


HEAT LOT 


HEAT LOT 


HEAT LOT 






9-11446 


9-11446 


9-10298 


0.2% Yield 










Avg (ksi) 


148.4 


165.8 


160.3 


152.6 


Std. Dev. (ksi) 


6.5 


0.2 


2.3 


0.6 


Ultimate Tensile 










Avg (ksi) 


188.5 


194.2 


192.0 


190.7 


Std. Dev. (ksi) 


4.1 


1.2 


2.4 


0.8 


Elongation 










Avg. (%) 


18.6 


18.7 


20.2 


13.0 


Std. Dev. 


1.6 


0.5 


0.6 





Reduction of Area 










Avg. (%) 


28.2 


30.3 


33.2 


38.0 


Std. Dev. 


2.5 


1.2 


1.5 





Charpy Impact Strength: 










Avg. (Ft-Lbs) 


19.8 


28.8 


29.3 


39.7 


Std. Dev. 


2.3 


1.0 


0.8 


2.9 



Grain Size (ASTM) 



5-8 



5-8 



6-8 



Chemical Data: C 


0.034 


0.027 


Ti 


0.98 


0.98 


S 


0.001 


0.002 


B 


<.0001 


<.001 


Fe 


17.672 


17.78 


Al 


.5 


0.510 


Cu 


.1 


0.06 


Ni 


53.5 


53.75 


Co 


0.18 


0.34 


B 


0.003 


0.003 


P 


0.01 


0.013 


Si 


0.14 


0.13 


Mn 


.1 


0.08 


Mo 


2.99 


2.98 


Cr 


18.4 


18.0 


Se 


<.0003 


<.O0O3 


Pb 


<.0001 


<.0001 


Cb+Ta 


5.29 


5.34 



0.027 


0.023 


0.98 


0.910 


0.002 


0.002 


<.001 


<0.00001 


17.78 


18.55 


0.51 


0.52 


0.06 


0.050 


53.75 


53.05 


0.34 


0.41 


0.003 


0.003 


0.013 


0.010 


0.13 


0.100 


0.08 


0.120 


2.98 


2.940 


18.0 


17.950 


<.0003 


<0.0003 


<.0001 


<0.0001 


5.34 


5.360 



- 237 - 



TABLE 4 

NAS 9-17496 FRANGIBLE NUT AND BOOSTER CARTRIDGE 

QUALIFICATION TEST MATRIX 



Test 

Room Temp. 
Firing 


Nut 

Group 

NO 

V 

I 


Functional 
Temp 

(-F) 

70-F 
70-F 


Preload 

Tension Mount 
(lbs) (in-lb) 

350,000 
270,000 


Booster 

Cartridges 

Dual/Single 

Single 
Single 


Functiona 
(pass/fail) 

Passed 
Passed 

Passed 
Passed 

Passed 
Passed 
Passed 

Passed 
Passed 
Passed 

Passed 

Passed 

Failed 
Failed 

Passed 
Passed 
Passed 


I 


High Temp. 
Firing 


II 
II 


+200-F 
+200-F 


270,000 
270,000 






Single 
Dual 




Low Temp. 
Firing 


III 
III 
III 


-65-F 
-65-F 
-65-F 


270,000 
270,000 
270,000 







Single 
Single 
Dual 




Low Temp. 
Firing with 
limit Axial Load 


V 

I 

V 


-65-F 
-65-F 
-65-F 


415,270 
415,270 
415,270 


53,725 
53,725 
53,725 


Single 

Dual 

Dual 




Room Temp. 
Firing with 
Zero Preload 


VI 


70-F 


No Load 





Single 




Margin Demo. 
Firings 


Vll 
Vll 
Vll 


70-F 
70-F 
70-F 


270,000 
270,000 
270,000 







Single 
Single 
Single 


(115% Web) 
(126% Web) 
(120% Web) 


85% Booster 
Cartridge 
Margin Demo. 
Firing 


IV 
IV 

Vlll 


70-F 
70-F 
70-F 


270,000 
270,000 
270,000 







Single 
Single 
Single 





- 238 - 



TABLE 5 
FAILURE INVESTIGATION TEST MATRIX 



Test 


Preload 


Nut Web 


Chamfered 


Booster 


Booster 


Results 






Thicknesses 


Outer Ledge 


Load 


Bore Area 






(Klbs) 


(%) 


(Y/N) 


(%) 


(%) 


(Pass/Fail) 


1 





100 


N 


110 


110 


Fail 


2 





100 


N 


115 


115 


Fail 


3 





100 


N 


120 


120 


Pass 


4 





80 


N 


100 


100 


Fail 


5 





100 


Y 


110 


110 


Pass 


6 


270 


100 


Y 


100 


100 


Fail 


7 





100 


N 


100 


120 


Fail 


8 





100 


Y 


100 


120 


Pass 


9 





100 


Y 


105 


105 


Fail 


10 


270 


115 


Y 


100 


120 


Pass 



TABLE 6 
NAS 9-17674 FRANGIBLE NUT AND BOOSTER CARTRIDGE QUALIFICATION TEST MATRIX 



Test 



Nut 

Group 

NO 



Functional 

Temp 

(-F) 



Pre-Load 
tension Mount 
(lbs) (in-lb) 



Booster 

Cartridges 

Dual/Single 



85% Booster G 70-F 270,000 Single 

Cartridge G 70-F 270,000 Single 

Margin Demo. G 70-F 270,000 Single 

Firing 

* -302 Nut represents nominal web thickness and chamfered outer ledges. 

** -102 Nut represents 115% nominal web thickness with chamfered outer ledges. 



Functional 
(pass/fail) 



Low Temp. 


C 


-65-F 


270,000 





Single 


Passed 




Firing 
















Low Temp. 


E 


-65-F 


415,270 


53,725 


Single 


Passed 




Firing with 


E 


-65-F 


415,270 


53,725 


Single 


Passed 




Limit Axial Load 


E 


-65-F 


415,270 


53,725 


Dual 


Passed 




Room Temp. 


D 


70-F 


No Load 





Single 


Failed 




Firing with 


D 


70-F 


No Load 





Single 


Failed 




Zero Preload 


D 


70-F 


No Load 





Single 


Passed 


(-302 Nut)* 




D 


70-F 


No Load 





Single 


Passed 


(-302 Nut)* 




D 


70-F 


No Load 





Single 


Passed 


(-302 Nut)* 


Margin Demo. 


G 


70-F 


270,000 





Single 


Failed 


(120% Web) 


Firings 


G 


70-F 


270,000 





Single 


Passed 


(-102 Nut)** 




G 


70-F 


270,000 





Single 


Passed 


(-102 Nut)** 



Passed 
Passed 
Passed 



- 239 - 




Figure 1. Illustration of orbiter/external tank aft attach 
interface and cross section of 2.5 inch frangible 
nut installation. 



ISOMICA 
DISKS 




1920 mg . 

RDX —I 



ISOMICA 
DISKS 



HOUSING 




1920 mg 
RDX 



HOUSING 



-401 CONFIGURATION 



-402 CONFIGURATION 



Figure 3. Illustration of 2.5-inch booster cartridge 

-401 configuration and modified design, 

-402 configuration. 




FRANGIBLE WEBS 



BOOSTER 
PORTS 



TOP VIEW OF 2.5-INCH 
FRANGIBLE NUT 




SECTION A-A 
FRANGIBLE NUT 
SEPARATION PLANE 



Figure 2. 2.5-inch frangible nut separation plane and 
frangible webs. 




-301 FRANGIBLE NUT WEBS 




115% (0.149") 
120% (0.156") 

L 



-1 01 , -1 02 MARGIN NUT WEBS 

Figure 4. 2.5-inch frangible nut nominal web thickness, 

(-301 configuration), 120% nominal web thickness, 

(-101 configuration) and 115% nominal web thickness 

(-102 configuration). 



- 240 - 



■■' ■ i r r-r- — r~^s^r * > 3 # g * W 







Figure 5. 100 X micrograph of original frangible nut 
supplier's Inconel 718. 





^ * ■ ... ■ 






Figure 7. 100X micrograph of Inconel 718 used 
inqualification test program under NASA contract 
NAS 9-17496. 



FRACTURED 
WEBS 



JL OUTER LEDGE 

GAP 




GAP DEVELOPED 
FROM SINGLE BOOSTER 
CARTRIDGE FIRING 



Figure 6. Illustration of clamshell motion experienced by 
2.5-inch frangible nut during single booster 
cartridge firing. 




TOP VIEW OF FRANGIBLE NUT 




SECTION A 

Figure 8. Illustration of material removal from outer 
ledge of 2.5-inch frangible nut. 



- 241 - 





Figure 9. 100X Micrograph of Inconel 718 Used in 
Production Lot under NASA Contract 
NAS 9-17496. 



Figure 10. 100X Micrograph of Inconel 718 Used in 
Qualification Lot under NASA Contract 
NAS 9-17674. 



- 242 - 






BOLT CUTTER FUNCTIONAL EVALUATION 

S. Goldstein, T. E. Wong, S. W. Frost, J. V. Gageby, and R. B. Pan 

The Aerospace Corporation 
2350 E. El Segundo Blvd., El Segundo CA 90245 



/ 



fi. 



t 



ABSTRACT 

The Aerospace Corporation has been implementing finite difference and finite 
element codes for the analysis of a variety of explosive ordnance devices. Both 
MESA-2D and DYNA3D have been used to evaluate the role of several design 
parameters on the performance of a satellite separation system bolt cutter. 
Due to a lack of high strain rate response data for the materials involved, the 
properties for the bolt cutter and the bolt were selected to achieve agreement 
between computer simulation and observed characteristics of the recovered 
test hardware. The calculations provided insight into design parameters such 
as the cutter blade kinetic energy, the preload on the bolt, the relative position 
of the anvil, and the anvil shape. Modeling of the cutting process clarifies 
metallographic observation of both cut and uncut bolts obtained from several 
tests. Understanding the physical processes involved in bolt cutter operation 
may suggest certain design modifications that could improve performance 
margin without increasing environmental shock response levels. 



BACKGROUND 

The Aerospace Corporation Explosive Ordnance 
Office (EOO) was given hardware from a series 
of satellite separation system ground tests 
wherein several bolt cutters successfully 
severed the interfacing bolts and others did 
not. EOO also obtained a severed bolt from a 
lot acceptance test of the cutter. The EOO 
was asked to assess causes of the anomalous 
performance and to determine corrective 
actions. A multi-disciplinary team was 
assembled and a review initiated. 

The bolt cutter used in this application was 
developed in 1 972 by Quantic (a.k.a. Whittaker 
or Holex) for McDonnell Douglas, Huntington 
Beach. A family of cutters known by part 
number R13200 has since evolved. The 
Quantic outline drawing for R 13200 states that 
its severance capacity is a 5/1 6 inch diameter- 
A286 CRES bolt having a tensile strength from 
180 to 210 ksi and tensioned between zero 
and 6000 lbs. A photograph of the hardware 
is shown in Figure 1. 

A finite element model, to be discussed further 



in a later section, is shown in Figure 2 and 
illustrates the configuration of the installed bolt 
and cutter before functioning. The bolt cutter 
consists of an explosive initiator, a chisel 
shaped cutter blade, a blade positioning shear 
pin, and an anvil in a cylindrical housing. The 
bolt to be cut fits through a clearance hole in 
the housing that places it against the anvil. 
When the initiator is functioned, the blade is 
accelerated and impacts the bolt. In both 
system separation and lot acceptance testing 
of the bolt cutter, it is seen that the cutter 
blade penetrates only part way through the bolt 
diameter. The cutting process is completed by 
fracture of the bolt. 

The initial finding of the team was that the 
ductility of the bolt used in the anomalous 
system separation tests was not compatible 
with the specification in the bolt cutter outline 
drawing. That is, a more ductile bolt than the 
R13200 bolt cutter requires had been used. 
The bolts used in the tests were solution 
annealed and aged to AMS 5737. This 
specification only requires a minimum ultimate 
tensile strength (UTS) of 140 ksi. 



Presented to the Second NASA/DOD/DOE Pyrotechnic Workshop, February 8-9, 1 994 



- 243 - 



To obtain the 180-210 ksi UTS, the A286 
material requires cold working per AMS 5731 . 
The bolts used in the system separation tests 
had not been cold worked. It was found that 
the Quantic target bolts, part number F12496, 
used in bolt cutter lot acceptance tests are cold 
worked. The F12496 drawing states that the 
target bolt material comply with AMS 5731 
and be cold worked to obtain 1 80-21 ksi UTS 
following heat treat per Mil-H-6875. Since 
1972, a large number of bolt cutters from 
many production lots have successfully severed 
the F12496 target bolts. No data base was 
found to assess cutter performance with A286 
bolts which had not been cold worked. 



METALLURGICAL ASSESSMENT OF A286 
BOLTS 

Metallurgical analyses were performed to infer 
the role of each parameter in the cutting 
process. The analyses were performed on 
fractured segments of a short bolt obtained 
from a Quantic lot acceptance test and on both 
fractured and unfractured long bolts from the 
system separation tests. The analyses included 
a microscopic examination of the fracture 
surfaces, metallurgical studies of the regions of 
deformation and fragmentation at the beginning 
of the cutting process, and measurements of 
material hardness and microstructure. 



The team also found that the mass of the bolts 
used in the system separation tests was at 
least six times greater than the Quantic test 
target bolt. The greater mass is due to greater 
length and end diameter, which is required for 
installation. The concern then was the lack of 
information on the effect of bolt inertia on bolt 
cutter performance. 

The team directed efforts toward analyzing the 
cut and uncut test bolts and in attempting to 
duplicate the cutter performance analytically. 
A F12496 target bolt, used in a bolt cutter lot 
acceptance test, was obtained from Quantic 
and also analyzed. In addition to assessing bolt 
inertia effects, the team attempted to 
determine the effect of bolt tension on the bolt 
cutter performance. The parameters assumed 
to affect the ability of the bolt cutter to sever 
the bolts are: 

• bolt configuration 

• ductility of the bolt material 

• preload in the bolt 

Other parameters such as gapping between the 
bolt and anvil and the anvil configuration were 
also considered. 

The following are the results of the material* 
analysis from metallographic evaluation of the 
test bolts, descriptions of the analytic modeling 
techniques, a comparison of model attributes 
and the team conclusions. 



These studies, and an examination of the 
photographs in Figures 3 through 9 lead to the 
following observations: 

• Grain size and hardness differences 
exist between the long and short bolts. 
The short bolt had a fine grain size and 
high hardness (Re 42), and was 
consistently severed. The long bolt 
was larger grained and softer (Re 35), 
and was not consistently severed. See 
Figures 10a and 10b. 

• The long and short bolts which had 
been successfully severed exhibited 
adiabatic shear bands in the deformed 
material regions adjacent to both the 
cutter blade and the anvil. Adiabatic 
shear bands are regions of highly 
localized plastic deformation resulting 
from the high material temperatures 
that are caused by high strain rate 
loading. 

• No evidence of adiabatic shear bands 
were seen in the deformed material 
regions adjacent to the anvil on the 
long bolt which had failed to separate. 



MODELING WITH MESA-2D 

MESA- 2D is a finite difference code that was 
used to analyze the behavior of the bolt cutter 
and bolt during the early time portion of its 
functioning. MESA is a reactive hydrodynamic 



- 244 - 



code that assumes, to a first approximation, 
that material behavior can be described by fluid 
dynamics when strong shocks are present. 
The equations of motion to be solved are then 
the time dependent nonlinear equations of 
motion for compressible fluids. 

Throughout a calculation, MESA-2D computes 
and records all relevant dynamic and 
thermodynamic properties for each cell (mass 
element) in the model. These variables could 
include position, velocity, pressure, internal 
energy, temperature, density, intrinsic sound 
speed, elastic and plastic work, plastic strain, 
strain rate, and deviator stress. All of these 
quantities output in graphical form or in tabular 
form for further analysis. By integrating over 
very small time steps, typically less than 1 
nanosec, the MESA calculation can handle 
impulsive loading of materials and allows their 
dynamics to be resolved with sufficient 
accuracy to elucidate the physical processes 
involved [1J. 

The numerical integrations required by the 
calculations were performed with coordinate 
meshes of between 40,000 and 60,000 cells. 
This gives better than 0.1 mm resolution, 
which is required to understand small systems 
such as the bolt cutter. 

Finite Difference Models 

Figure 11 shows the cutter blade, the anvil, 
and the bolt to be cut that were included in the 
MESA finite difference model. It also shows 
the particle velocity distribution after 20/ssec. 
An alternative configuration was also 
developed in which the massive ends of the 
flight bolt were eliminated. The models were 
analyzed using slab geometry and transmissive 
boundary conditions for the hydrodynamic 
equations. Bolt preload was not included. 

In the computer simulations, the available 
material properties of 304 stainless steel were- 
used as the basis for the properties of all 
metallic components. The yield strength and 
shear modulus were adjusted by using 4340 
steel for the blade and A286 for the bolt. 
Strain rate effects were accounted for by 
allowing these constants to vary [2, 3, 4). 



The simulations were assumed to start when 
the cutter blade begins to move, and neglects 
the functioning of the initiator and the transfer 
of energy to the cutter blade. 

The initiator consists of 70% by weight 
ammonium perchlorate <AP), 27% 
polybutadiene acrylic acid (PBAA), and 3% 
combined zirconium barium peroxide (ZBP), 
ferrix oxide (FO), and epoxy resin. The ZBP 
and FO as oxidizers will enhance the 
performance of the main constituent, AP. The 
remaining materials are inert binders. The 
material properties of all these materials are not 
known. An upper and lower bound estimate of 
kinetic energy output was made from the 
available chemical energy of the AP assuming 
instantaneous energy release via detonation. 
Since the exact energy partition is unknown, 
trial computer models were run using several 
candidate velocities within these limits. The 
cutter velocity was determined by trial and 
error, matching the resulting penetration into 
the bolt to the experimental data. This velocity 
was 332 m/sec. 

Computational Results 

The calculations began at time zero with the 
bolt cutter blade poised to impact the surface 
of the bolt and with the initial constant velocity 
of 332 m/sec. The proper penetration of the 
cutter blade into the material to agree with the 
test data from Figures 5, 6 and 7 is achieved in 
approximately 20 //sec. The material interfaces 
show that both light and heavy bolts have 
responded identically to this point. The time 
elapsed to this point in the cutting process is 
one order of magnitude smaller than the time it 
had previously been assumed by the 
community for bolt cutter function. 

Compression and tension waves propagating 
through the bolt show that there is no net 
motion of the bolt ends since the particle 
velocities of the end cells go to zero. The 
velocity of the cutter blade also changes 
direction several times after it penetrates the 
bolt. This is evidence of an oscillation that is 
set up which will cause the blade to bounce 
back. 



- 245 - 



The model shows, as does the test hardware, 
that all material deformation occurs within a 1 
cm radius of the impact point of the cutter 
blade. No rigid body motion of the bolt is 
required for penetration, and indeed the 
coordinates of the bolt ends do not change 
throughout the cutting process in the 
calculation. 

Shear deformation can be discerned from 
inflections in the particle velocity distributions 
as early as 10 //sec. These patterns are 
apparent in Figure 11. This result agrees with 
the evidence of the same behavior in the 
photomicrographs of the cut surfaces. Ejection 
of particles from the top surface of the bolt can 
be seen. There is also a small crack that 
appears near the tip of the cutter blade. All of 
these features were seen in the hardware, 
especially Figures 5 and 6. The indentation 
from the anvil on the underside of the bolt can 
also be seen beginning to form although it is 
not obvious until some time later. 

The material deformation is due almost entirely 
to plastic work. The elastic contribution is 
between two and three orders of magnitude 
smaller than the plastic, and the penetration 
process is completed before the effects of any 
elastic waves can be seen. Therefore, the ends 
of the bolt, and whether or not they are 
massive, may have no effect on cutter 
performance. 

Blunting of the cutter blade edge occurs as 
well. Figures 12 and 13 show this in the 
hardware. The assumption had been that this 
blunting resulted from the impact of the blade 
against the anvil after the bolt had separated. 
While there is undoubtedly some effect from 
this, the blade edge is also blunted by erosion 
during the bolt penetration process. 

As configured, the MESA calculations do not 
show that the cutter completely severs the 
bolt. This may be partly attributed to- 
insufficient brittleness in the material 
description. However, the highest shear 
locations match those in the photomicrographs 
of the test bolt that failed to cut. The shear 
deformation regions that originate at the anvil 
side of the bolt are not as clear with the 



resolution available in the calculation. 

Additional calculations were performed using 
both smaller and larger velocities for the cutter 
blade. When the velocity is 133 m/sec, the 
blade does not penetrate far enough to match 
the data. When it is increased to 431 m/sec, 
it is possible to separate the bolt by penetration 
alone, independent of the formation of a shear 
fracture. These calculations indicated that 
depth of penetration of the blade into the bolt 
is dependent on this variable alone. This is 
consistent with the results of various empirical 
penetration analyses for projectiles [5], which 
show that penetration depth is a function of 
the velocity of the penetrator and the ratio of 
densities of the materials of penetrator and 
target. Since both blade and bolt have the 
same density, the only other determining factor 
is penetrator velocity. 



MODELING WITH DYNA3D 

Due to analytical considerations of nonlinear 
dynamics and stress wave propagation effects 
in bolt cutter structural response, transient 
dynamic analyses were performed using the 
DYNA3D code. The DYNA3D code is an 
explicit, nonlinear, finite element analysis code 
developed by Lawrence Livermore National 
Laboratory [8]. It has a sophisticated 
simulation capability for handling frictional and 
sliding interactions between independent 
bodies. 

A built-in feature in the DYNA3D code was 
selected for modeling the sliding surface failure 
behavior. A failure criterion based on the total 
cumulative effective plastic strain is defined for 
the elements adjacent to the contact surface. 
When the rate-dependent plastic strain value 
within an element satisfies this failure criteria, 
the element is removed from further calculation 
and a new sliding surface boundary is defined. 

Parameters compared in this study include the 
approaching speed of the cutter blade, the 
applied bolt preload, the gap clearance between 
the bolt and the anvil, and the contact surface 
area of the anvil. TABLE I lists the four 
parameters and their variations considered in 



- 246 - 



the finite element analysis matrix. In this table, 
the 3,270 lbs bolt preload and a maximum gap 
allowable of 0.065 inch between the bolt and 
anvil were based on drawing specifications. 
The loss of preload and a 50% reduction in 
anvil's contact area were chosen to study their 
impact in cutter performance. 

The speed for the blade was determined using 
system separation test data. In these tests, six 
A286 bolts were preloaded to 3,270 lbs. Two 
of the bolts had a cutting depth of 40% of the 
bolt diameter and did not fracture. The other 
four bolts had a slightly deeper blade 
penetration, were totally severed and the cutter 
blades also indented the anvil. The transient 
dynamic analysis duplicated these conditions 
by using a 6,000 in/sec (152 m/sec) speed. 

Finite Element Analysis Matrix 

The Taguchi experimental design technique [6, 
7] was then adopted for establishing the 
analysis matrix. This technique was also used 
to analyze the finite element calculation results 
to identify the optimum bolt cutter 
configuration, especially in relation to preload 
or applied tension on the bolt. 

A Taguchi analysis matrix with 8 study cases 
(a L 8 orthogonal array [6, 7]), shown in TABLE 
II, was established to evaluate the criticality of 
the four chosen parameters mentioned above. 
Interaction effects between these parameters 
were assumed to be negligible. Based on the 
analysis matrix and the chosen parameter listed 
in TABLE II, 8 different finite element models 
were constructed. A baseline finite element 
model of the bolt cutter configuration (case 1 
in TABLE II) as shown in Figure 2. Due to 
symmetry of the bolt cutter geometry, only one 
half of the bolt cutter configuration was 
modeled. This model consists of 618 solid 
elements and 1027 nodes to simulate the 60 
degree cutter blade, the 5/16 inch diameter 
bolt, and the anvil. 

Transient D ynamic Analysis 

TABLE III lists the mechanical properties used 
in the analysis. The values chosen for S7 tool 
steel and A286 stainless steel were obtained 



from references [9] and 110], respectively. 

In the finite element analysis model, 
nonreflecting boundaries were assumed at the 
two ends of the bolt and the anvil to prevent 
artificial stress wave reflections re-entering the 
model and contaminating the results. A fixed 
end boundary condition was assumed at the 
lower end of the anvil. The bolt preload was 
first generated by applying pressure loading on 
the bolt with a built in dynamic relaxation 
option. The cutter blade with the appropriate 
approaching speed was then applied. A 
transient dynamic analysis was performed to 
estimate the damage in the bolt. The analysis 
results for the eight study cases are listed in 
the last column of TABLE II. The 0's and 1's 
are corresponding to a partial or a complete 
cutting of the bolt, respectively. Figure 14 
shows the simulation results at 0.4 ms for the 
baseline model in TABLE II. In Figure 15, with 
a finer mesh model, it can be seen that the bolt 
is completely severed by the cutter. It can be 
seen that the failure configuration matches 
fairly well with the test data in Figure 3, 

Bolt-Cutter Performance Assessment 

From finite element analysis results listed in the 
last column of TABLE II, the bolt cutter 
performance, based on variation levels for each 
parameter, can be summarized. For example, 
the cutter performed well with a bolt preload 
(parameter A) of 3,270 lbs (level 1 ). It resulted 
in three successful cuts and one failure. The 
cutter performed poorly with parameter A at 
level 2 (zero preload), with only one success 
and three failures. Therefore, the analysis 
results of level sum A, and A 2 are: 



A,= 1 
A 2 = 1 



+ 1+1+0 = 3 cuts 
+ + + = 1 failure 



Total 



= 4 cases 



Other parameter sums are similarly calculated 
and summarized in TABLE IV. These results are 
also plotted in Figure 16 in bar chart format in 
terms of the percentage of success. It can be 
seen that the bolt cutter performance can be 
improved with the design parameter setting of 
A„ B 1# C 1# D 2 . That is, it is desirable to 



- 247 - 



improve the cutter performance by applying a 
bolt preload of 3,270 lb, providing sufficient 
energy for the cutter blade to reach a speed of 
6,000 in/sec, ensuring that no gap exists 
between the bolt and the anvil before firing, 
and by reducing the anvil and the bolt contact 
surface area by half. 

Omeaa Transformation 

In order to verify the assumption that 
interaction effects between these four 
parameters are negligible and to estimate the 
response of the optimum condition of the bolt 
cutter system, the omega transformation 
technique [7] can be used. The omega 
transformation is defined as follows: 

Q(P) - -10log(1/P-1)dB 

where P is the percentage of success. For the 
cases of cutter failure (0%) and cutter success 
(100%), they are treated as follows: 

0% case - Consider this as 1 /(number of cases) 
and perform the omega transformation. For the 
current study, the number of cases is 8; thus, 
(1/8)x100 = 12.5% orQ<12.5%)=-8.45 dB. 

100% case - Consider this as (number of cases 
- 1 )/(number of cases) and perform the omega 
transformation. For the current study, 
K8-1)/8]x100 = 87.5% orQ<87.5%} = 8.45 dB. 

Based on the approach as shown in [7], the 
optimum response, m, can be estimated by an 
additive model. 

mtA^^Dj) = T A1 + T B1 + T C1 + T 02 - 3 x T 
= Q(75%) + Q(75%) + Q(75%) 

+ 0(75%) - 3 x Q(50%) 
= 4.77 + 4.77 + 4.77 + 4.77- 

3x0 
= 19.08 dB ( > 8.45 dB) 
= 98.8% ( > 87.5% ). 

Here T is the overall mean percentage of 
success for all cases analyzed in Table IV and 
T Xy is the average for parameter X at level y. 
Thus, under the optimum conditions, the bolt 
could be totally severed by the design of 
A 1 B 1 C 1 D 2 . This was later confirmed by the 



finite element analysis prediction. This result 
indicates that the additive model is adequate 
for describing the dependence of the structural 
response on various parameters, and also 
confirms that the assumption of negligible 
interaction effects between various governing 
parameters was valid. 



CONCLUSIONS 

The evidence obtained from the metallurgical 
examination of test hardware suggests that 
bolt severance from the impact of the bolt 
cutter blade occurs as a result of a combination 
of processes. These include: 



reduction of the bolt diameter 
penetration of the blade; 



by 



reduction of the bolt diameter by 
penetration of the anvil; 

wedge opening forces generated by the 
cutter blade as it penetrates; 

applied preload on the bolt; and 

adiabatic shear band formation under 
the combined effects of shock heating 
and the applied stresses on the bolt. 



The two analysis techniques that were 
employed proved to be complementary to each 
other in that they each were able to elucidate 
different features of the ductile bolt behavior 
and of the governing design parameters of the 
system. In addition to the above conclusions, 
which they confirm, are the following. 

The MESA-2D analysis indicates: 

• The bolt cutting process is completed 
in less than 60 //sec. 

• Neither the length nor mass of the bolt 
has any effect on the ability of the 
cutter blade to penetrate the bolt. 

• The depth of penetration of the cutter 
blade into the bolt is a function of the 

cutter blade velocity. 



- 248 - 



• The bolt fractures due to shear in the 
second part of the separation process. 

The DYNA3D analysis also indicates: 

• With the available explosive energy, a 
preload in the bolt is necessary for the 
fracture to occur and complete the bolt 
separation. 

• For effective severing, there should be 
contact between the bottom surface of 
the bolt and the anvil. 



REFERENCES 



[1] S. T. Bennion and S. P. Clancy, 
"MESA-2D (Version 4)", Los Alamos 
National Laboratory, LANL Report LA- 
CP-91-173, 1991. 

[2] E. L. Lee, H. C. Hornig, and J. W. Kury, 
"Adiabatic Expansion of High Explosive 
Detonation Products", Lawrence 
Livermore National Laboratory, LLNL 
Report UCRL-50422, 1 968. 



• The bolt cutter is more effective if the 

surface area of the anvil in contact 
with the bolt is decreased. 

This last conclusion presents a possible design 
modification that is an alternative to increasing 
the kinetic energy of the cutter blade with 
additional explosive. Increasing the amount of 
explosive could increase the shock from 
functioning the device, whereas a change in 
anvil configuration would not. 

Further work is still needed on the bolt cutter 
system to analyze the performance under 
conditions where the cutter blade has a 
non-parallel impact to the cross section of the 
bolt. 



ACKNOWLEDGMENTS 

The authors would like to thank the following 
individuals for their contributions to this work: 

G. Wade of Quantic Industries, for providing 
drawings, hardware, and other information on 
the design and materials in the 13200 bolt 
cutter; A. M. Boyajian, The Aerospace 
Corporation program office, for his support of 
this work; R. W. Postma, L. Gurevich, G. T. 
Ikeda and R. M. Macheske, The Aerospace 
Corporation, and J . Yokum of Defense- 
Systems, Inc. for their interest and participation 
in many valuable technical discussions. 



[3] D. J. Steinberg, S. G. Cochran, and M. 
W. Guinan, "A Constitutive Model for 
Metals Applicable to High Strain Rate", 
J. AppL Phys. 51 (3), 1498 (1980). 

[4] G. R. Johnson and W. H. Cook, 
" Fracture Characteristics of Three 
Metals Subjected to Various Strains, 
Strain Rates, Temperatures and 
Pressures", Eng. Frac. Mech. 21 (1), 
31 (1985). 

[5] Joint Technical Coordinating Group for 
Munitions Effectiveness (Anti-Air), 
Aerial Target Vulnerability Subgroup, 
Penetration Equations Handbook for 
Kinetic Energy Penetrators (U), 61 
JTCG/ME-77-1 6 Rev. 1,15 Oct. 1 985. 

[6] Phadke, M. S., Quality Engineering 
Using Robust Design . Prentice Hall, 
Englewood Cliffs, NJ, 1989. 

[7] Mori, T., The New Experimental 
Design , ASI Press, 1990. 

[8] Whirley, R. G. and Hallquist, J. 0., 
"DYNA3D Users Manual," Lawrence 
Livermore Laboratory, Rept. UCRL-MA- 
107254, May 1991. 

[9] American Society for Metals, Metals 
Handbook , Vol. 3, 9th Edition, 1980. 

[10] Frost, S. W., "Metallurgical Evaluation 
of Separation Bolt, " Aerospace 
Corporation Interoffice 
Correspondence, 9 Nov. 1992. 



- 249 - 



Table I. Parameters for Bolt Cutter Study 



Parameter 


Description 


Variation Level 


1 


2 


A 


Bolt Preload, lbs. 


3270 





B 


Cutler Blade Speed in./sec. 


6000 


5000 


C 


Gap between bolt and Anvil, 
in. 





0.065 


D 


Anvil Surface Reduction, % 





50 



Table II. Finite Element Analysis Matrix 
(Taguchi Lg Orthogonal Array) 



Analysis Run 


Parameters 


Results * 


A 


B 


C 


D 


1 


1 


1 


1 


1 


1 


2 


1 


1 


2 


2 


1 


3 


1 


2 


1 


2 


1 


4 


1 


2 


2 


1 





5 


2 


1 


1 


2 


1 


6 


2 


1 


2 


1 





7 


2 


2 


1 


1 





8 


2 


2 


2 


2 






* 1 represents bolt totally fractured, represents bolt partially fractured 



- 250 - 



Table III. Mechanical Properties Used for DYNA3D Analysis 



Structure 


Material 
Type 


Poisson's 
Ratio 


Yield 

Strength 

(ksi) 


Tensile 

Strength 

(ksi) 


Elongation 
(%) 


Cutter 
Blade 


S7 Tool 
Steel 


0.31 


210 


315 


7 


Bolt 


A286 

Stainless 

Steel 


0.31 


139 


172 


20 


Anvil 


A286 

Stainless 

Steel 


0.31 


139 


172 


20 



Table IV. Cutting Efficiency 



Parameter 


Variation Level 


1 


2 


A 


3(75) 


1 (25) 


B 


3(75) 


1 (25) 


C 


3(75) 


1 (25) ... 


D 


1 (25) 


3(75) 



Numbers in parentheses are percentage (%). 



- 251 - 





(iMlji|!|Hi|||l 



liiliilll 



Figure 1 : Top photo is the configuration of the long bolt used in system separation 
tests. One bolt is fully separated, the other is not. Bottom photo shows the 
bolt cutter with the internal blade visible through the hole at the anvil end 
of the cutter. 



- 252 - 



l£&&i 



I 
i 




Figure 2: Bolt cutter and bolt finite element model showing 1/2 the configuration 
before functioning. 





Figure 3: 

Top photo shows separation 
area on the short bolt. Bottom 
photo is an SEM micrograph 
providing a face-on view of the 
fracture on the top right. Note 
secondary cracks on the blade 
cut area and complex fracture 
surface in lower quadrants. 



- 254 - 




■ ; Kk 



Pill: ^?^IS; 




Figure 4: 

Top photo shows separation 
area on the long bolt. Bottom 
photo is an oblique view of the 
fracture seen on the top left. 
Note the limited extent of the 
blade cut surface remaining on 
this segment. 



- 255 - 








Figure 5: Top photo shows impact area of the long bolt which failed to separate. 
Bottom photo shows impact area after polishing away 1/4 of the bolt 
diameter from the side of the bolt. Note the small 45° crack at the tip of the 
notch produced by the blade impact. 



- 256 - 






Figure 6: Top photo shows impact area of the long bolt after polishing away 1/3 
of the bolt diameter. Bottom photo shows same area after etching (with 
dilute hydrochloric and nitric acid mixture). Note in this section there is 
no crack at the tip of the notch. Adiabatic shear bands of localized 
deformation are seen on both sides of the notch. 



- 257 - 





Figure 7: Top photo shows impact area of the long bolt after polishing away 1/2 
of the bolt diameter. Bottom photo shows same area after etching. 
Adiabatic shear bands emerge from the sides of the notch. There is no 
evidence of similar bands adjacent to the anvil. 



- 258 - 





Figure 8: Top photo shows separation area of the short bolt. Bottom photo is a view 
of the midplane of the bolt segment shown on the top left. Note that a band 
of localized deformation (adiabatic shear band) has formed in the deformed 
material above the anvil (see arrow). 



- 259 - 




Figure 9: 

Top photo shows polished and 
etched section through the 
deformed area above the anvil 
on the short bolt. Bottom 
photo shows polished and 
etched section through the 
deformed area above the anvil 
on the fractured long bolt. 
Adiabatic shear bands of 
localized deformation are 
emerging in the area of the bolt 
near the corner of the anvil. 



50x 




200x 



260 - 



'\ 



/-> 



;«**%. 



X 



\ --^P* 






r\. 



\ 



V 



^ 
^ 



- : s»: 









ViiJk. 



\3 



1000X 



Figure 10a: Optical micrographs showing typical microstructure in the center of a 
transverse section through the long bolt (etched with Glyceregia). 
Microhardness measurements indicate Rockwell "C" of 35-36. This 
corresponds to a tensile strength of 160 ksi. 



■• TS-v"« ?■*':*<. --?-'- 











><2& ; - ^ v . ^ V v*s*nJ^a *> -* 




1000X 



Figure 10b: Optical micrographs showing typical microstructure in the center of a 
transverse section through the short bolt (etched with Glyceregia). 
Microhardness measurements indicate Rockwell M C H of 41-43. This 
corresponds to a tensile strength of 192 ksi. 



- 261 - 



VECTOR VELOCITIES 



TIME 



20.0084 



LOCAL MAX 2.857E-02 
^ 



GLOBAL MAX 2 . 857E-02 



SCALED VEC 2.857E-02 



i 

CD 

I 



O 
DC 



12.0 _ 



10.0 _ 



8.0 _ 



6.0 _ 



4.0 - 



2.0 



0.0 



- 








r -■■■ ^ 




~ 


[ t 1 


l t i I 






t I I 


i i i \ 






1 f 1 
1 1 I 


mm 
MM 




— 


I 1 1 


MM 






t t I 


ml 




— 


t 1 I 


i i i i 






1 I t 


l M I 






1 t I 
I 1 I 


till 

I I I \ 




\l I 


III/ 


k i\ i i i i i A 77 


>v\ l l K l>&7 


- . 






- I 






























Z »-«.>*»»•«. — ZIZ % 


,p^~~ + * + + + *< 
1^-** + + + + + * 

, /L _ — «• ^ >» *• * * » 
#» ^ - ^ + * + * * - 
































{* 


^\ 




"*? 


1| 


. ! 


— 






I i I i I 


,: : : 






} 1 1 I.I 1 



-6.5 -5.5 



-4.5 -3.5 



-2.5 -1.5 -.5 .5 
Z (CM) 



1.5 2.5 3.5 4.5 



5.5 



6.5 



Figure 1 1 . MESA-2D model of cutter blade and bolt 20 ixsec after impact. 





Figure 12: Top photo shows the cutting edge of the blade used in separating the long 
bolt. Bottom photo shows damage to the anvil resulting from impact of the 
blade after bolt separation. 



- 263 - 





Figure 1 3: Comparison of damage to the cutter blades used in separating the long and 
short bolts. Blade used for the short bolt is missing more material (bottom 
of each photo). 



- 264 - 







Figure 14: Bolt during impact of cutter blade at 0.4 msec in DYNA3D. 



- 265 - 





L. 



Figure 15: Fractured bolt configuration for finite element analysis. 



- 266 - 



100 




3270 

Bolt Preload (lb) 



6000 5000 
Cutter Blade Speed (in. /sec.) 



i 

CD 

i 




0.065 

Gap between Bolt & Anvil (in.) 



50 

Anvil Surface Reduction (%) 



Figure 16: Parameter effects in bolt fracture study. 



Choked Flow Effects in the NSI Driven Pin Puller* 

Keith A. GonthierWd Joseph M. Powers* 

Department of Aerospace and Mechanical Engineering 

University of Notre Dame 

Notre Dame, Indiana 46556-5637 

Abstract 

This paper presents an analysis for pyrotechnic combustion and pin motion in the NASA 
Standard Initiator (NSI) actuated pin puller. The conservation principles and constitutive 
relations for a multi-phase system are posed and reduced to a set of eight ordinary differential 
equations which are solved to predict the system performance. The model tracks the inter- 
actions of the unreacted, incompressible solid pyrotechnic, incompressible condensed phase 
combustion products, and gas phase combustion products. The model accounts for multi- 
ple pyrotechnic grains, variable burn surface area, and combustion product mass flow rates 
through an orifice located within the device. Pressure-time predictions compare favorably 
with experimental data. Results showing model sensitivity to changes in the cross- sectional 
area of the orifice are presented. 

Introduction 

Pyrotechnically actuated devices are widely used for aerospace applications. Examples 
of such devices are pin pullers, exploding bolts, and cable cutters. Full-scale modeling efforts 
of pyrotechnically driven systems are hindered by many complexities: three dimensional- 
ity, time-dependency, complex reaction kinetics, etc. Consequently, simple models have 
been the preferred choice of many researchers. x » 2 ' 3 ' 4 These models require that a number 
of assumptions be made; typically, a well stirred reactor is simulated; also, the combustion 
product composition is typically predicted using principles of equilibrium thermochemistry, 
and the combustion rate is modeled by a simple empirical expression. 

Recently, Gonthier and Powers 5 described a methodology for modeling pyrotechnic com- 
bustion driven systems which is based upon principles of mixture theory. Though this ap- 
proach still requires that simplifying assumptions be made, it offers a rational framework for 
1) accounting for systems in which unreacted solids and condensed phase products form a 
large fraction of the mass and volume of the total system, and 2) accounting for the transfer 
of mass, momentum, and energy both within and between phases. The methodology was 
illustrated by applying it to a device which is well characterized by experiments: the NSI 
driven pin puller. 



'Presented at the Second NASA Aerospace Pyrotechnic Systems Workshop, February 8-9, 1994, Sandia 
National Laboratories, Albuquerque, New Mexico. This study is supported by the NASA Lewis Research 
Center under Contract Number NAG-1335. Dr. Robert M. Stubbs is the contract monitor. 

* Graduate Research Assistant. 

* Assistant Professor, corresponding author. 



2 2-2% 
?G02 



/(o 



_ 2 6 9 . 1page134Lintem 



it? r-.M -v r 



The focus of this paper is on using the methodology presented in Ref. 5 to formulate a 
pin puller model which additionally accounts for the flow of combustion products through 
an orifice located within the device; the model is then used to determine the influence of 
product mass flow rates on the performance of the device. The present model also accounts 
for multiple pyrotechnic grains and variable burn surface area. The model presented in this 
paper is an extension of the model presented in Ref. 5 which did not account for product 
flow through the orifice, multiple grains, or variable burn surface area. 

Figure 1 depicts a cross-section of the NSI driven pin puller in its unretracted state. 6 
The primary pin, which will be referred to as the pin for the remainder of the paper, is 
driven by gases generated by the combustion of a pyrotechnic which is contained within the 
NSI assembly. Two NSFs are tightly threaded into the device's main body. Only one NSI 
need operate for the proper functioning of the pin puller; the second is a safety precaution in 
the event of failure of the first. The pyrotechnic consists of a 114 rag mixture of zirconium 
fuel (54.7 mg Zr) and potassium perchlorate oxidizer (59.3 mg KC10 A ). Initially a thin 
diaphragm tightly encloses the pyrotechnic. Combustion is initiated by the transfer of heat 
from an electric bridgewire to the pyrotechnic. Upon ignition, the pyrotechnic undergoes 
rapid chemical reaction producing both condensed phase and gas phase products. The 
high pressure products accelerate the combustion rate, burst the confining diaphragm, vent 
through the NSI port (labeled "port" in Fig. 1), and enter into the gas expansion chamber. 
Once in the chamber, the high pressure gas first causes a set of shear pins to fail, then pushes 
the pin. After the pin is stopped by crushing an energy absorbing cup, the operation of the 
device is complete. Peak pressures within the expansion chamber are typically around 50.0 
MPa; completion of the stroke requires approximately 0.5 ms. 6 

For sufficiently high NSI assembly/gas expansion chamber pressure ratios (~ 2.0), and 
for a fixed cross-sectional area of the NSI port, there exists a maximum flow rate of combus- 
tion product mass through the port. The occurrence of this maximum flow rate is referred 
to as a choked flow condition. Such a condition results in the maximum flux of energy 
into the expansion chamber; the energy contained within the chamber can then be used to 
perform work in moving the pin and can be lost to the surroundings in the form of heat. 
However, if the time scales associated with the flux of energy into the expansion chamber 
and the rate of heat lost from the products within the chamber to the surroundings are of 
the same magnitude, there may be insufficient energy available to move the pin; functional 
failure of the device would result. Therefore, it is possible that variations in the flow rate 
of product mass through the port may significantly affect the performance of the device. 

Included in this report are 1) a description of the model including both the formula- 
tion of the model in terms of the mass, momentum, and energy principles supplemented 
by geometrical and constitutive relations and the mathematical reductions used to refine 
the model into a form suitable for numerical computations, 2) model predictions and com- 
parisons with experimental results, and 3) results showing the sensitivity of the model to 
changes in the cross-sectional area of the NSI port. 

Model Description 

Assumptions for the model are as follows. As depicted in Fig. 2, the total system 
is taken to consist of three subsystems: incompressible solid pyrotechnic reactants (s), 
incompressible condensed phase products (cp), and gas phase products (p). The solid 
pyrotechnic is assumed to consist of N spherical grains having uniform instantaneous radii. 
The surroundings are taken to consist of the walls of the NSI assembly, the NSI port, and 

- 270 - 



1/4" 



stroke 



9/16" 



ZrKC10 4 

pyrotechnic 



expansion pon^ 
chamber. 




NASA Standard 
Initiator (NSI) 
assembly 



energy 

absorbing 

cup 



Figure 1: Cross-sectional view of the NSI driven pin puller. 



NSI port- 




Gas Expansion Chamber (2) — ^ 



combustion 
products 



^^s!H^^^ 



NSI Assembly (1) 

gas phase products (g) 
condensed phase products (cp) 





shear pin 
system boundary 



Figure 2: Schematic of the two component system for modeling choked flow effects. 

the gas expansion chamber. Both the NSI assembly and the gas expansion chamber are 
modeled as isothermal cylindrical vessels. The gas expansion chamber is bounded at one 
end by a movable, frictionless, adiabatic pin while the volume of the NSI assembly remains 
constant for all time. The NSI port is assumed to have zero volume and is characterized by 
its cross- sectional area. 

Mass and heat exchange between subsystems is allowed such that 1) mass can be trans- 
ferred from the solid pyrotechnic to both the condensed phase and gas phase products, 
and 2) heat can be transferred from the condensed phase to the gas phase products. The 
condensed phase - gas phase heat transfer rate is assumed to be sufficiently large such that 
thermal equilibrium between the product subsystems exists. There is no mass exchange 
between the system and the surroundings. Both product subsystems are allowed to interact 
across the system boundary in the form of heat exchanges. The gas phase products are al- 
lowed to do expansion work on the surroundings. No work exchange between subsystems is 



- 271 - 



allowed. Spatial variations within subsystems are neglected; consequently, all variables are 
only time-dependent and the total system is modeled as a well-stirred reactor. The kinetic 
energy of the subsystems is ignored, while an accounting is made of the kinetic energy of 
the bounding pin. Body forces are neglected. 

The rate of mass exchange from the reactant subsystem to the product subsystems is 
taken to be related to the gas phase pressure within the NSI assembly, namely dr/dt — 
—bP gi , where r is the instantaneous radii of the pyrotechnic grains, t is time, P gi is the 
gas phase pressure within the NSI assembly, and b and n are experimentally determined 
constants. All combustion is restricted to the burn surface of the pyrotechnic grains. In 
the absence of burn rate data for Zr/KClO^ we have chosen values for b and n so that 
pressure-time predictions of our model agree with experimental data. The equilibrium ther- 
mochemistry code CET89 7 calculated for the constant volume complete combustion of the 
Zrj KCIO4 mixture is used to predict the product composition; the initial total volume 
of the pin puller (0.95 cm 3 ) was used in this calculation since a significant portion of the 
system mass is contained within the gas expansion chamber at the time of complete com- 
bustion. The component gases are taken to be ideal with temperature-dependent specific 
heats. The specific heats are in the form of fourth-order polynomial curve fits given by the 
CET89 code and are not repeated here. 

The rate of gas phase product mass flowing from the NSI assembly, through the NSI port, 
and into the gas expansion chamber is modeled using standard principles of gas dynamics. 
The flow of condensed phase product mass through the port is assumed to be proportional 
to the gas phase mass flow rate. The only energy interaction between the NSI assembly and 
the gas expansion chamber is due to the energy flux associated with the exchange of mass 
between these two components. 

Using principles of mixture theory, a set of mass and energy evolution equations can be 
written for each subsystem contained within the NSI assembly and gas expansion chamber. 
These equations, coupled with an equation of motion for the pin, form a set of ordinary 
differential equations (ODE's) given by the following: 

fo(p»iV*i) = -PsxMn, (1) 

-77 (Pc Pl V cpi ) = ricpp^Abn - ro cp , (2) 

J t (Pgi v 9i) = C 1 - Vcp)ps 1 A b r b - ro„ (3) 

ft(P*i V *i e »i) " ~P*i e *i A b r b, ( 4 ) 

-77 (pc P iVc Pl e cpi ) = r] cp p Sl € 3l A h r b - h cpi m cp - Q cPi9l + Qc Pl , (5) 

J t (Pgi v 9i e 9i) = C 1 " nc P )ps l e 5l A b r b - h gi m g + Q cp , 9l + Q gi , (6) 

fa(pC P2 V C P2 ) = ™>CJ>, (7) 

J t (P82 V 9%) = ™S> ( 8 ) 

"77 \Pcj>2 *c P 2 e c P 2 ) == "cpi^cp — Qc P} g2 * *4cp2i \y) 

- 272 - 



^ (pg 3 V g2 e g2 ) = h gi m g + Q cp , g2 + Q g2 - W out2 , (10) 

™ P ^(z P ) = F P - (11) 

In these equations, the notation subscript "1" and subscript "2" are used to label quantities 
associated with the NSI and the gas expansion chamber, respectively. Subscripts "s", "cp", 
and u g" are used to label quantities associated with the solid pyrotechnic, condensed phase 
products, and gas phase products, respectively. The independent variable in Eqs. (1-11) is 
time t. Dependent variables are as follows: the density p 9i (here, and for the remainder of 
this report, the index i = 1,2 will be used to denote quantities associated with the NSI and 
the gas expansion chamber, respectively); the volumes V Sl , F cPt , V gi ; the specific internal 
energies e Bl , e cp ., e gi ; the specific enthalpies h cpi , h gi ; the pin position z v \ the pyrotechnic 
burn rate r&; the area of the burn surface Ab\ the rates of product mass flowing through the 
NSI port m cp , rh g ; the rates of heat transfer from the surroundings to the gas phase products 
Qim J the rates of heat transfer from the condensed phase products to the gas phase products 
Q C p,9i\ the rate of work done by product gases contained within the expansion chamber in 
moving the pin W ou t 2 ; an( i the net force on the pin F p . 

Constant parameters contained in Eqs. (1-11) are the mass of the pin m p , the density of 
the unreacted solid pyrotechnic p 5l , the density of the condensed phase products p cpi and 
/> cp2 , and the mass fraction of the products which are in the condensed phase r) cp . As it is 
understood that the pyrotechnic is contained entirely within the NSI, the notation subscript 
"1" will be dropped when referring to quantities associated with the solid pyrotechnic. Also, 
since p cpi = p cp2 = constant, these two quantities will be referred to as p cp . 

Equations (1-3) govern the evolution of mass and Eqs. (4-6) govern the evolution of 
energy for the solid pyrotechnic, the condensed phase products, and the gas phase products 
contained within the NSI, respectively. Equations (7) and (8) govern the evolution of mass 
and Eqs. (9) and (10) govern the evolution of energy for the condensed phase products and 
gas phase products contained within the gas expansion chamber, respectively. Equation 
(11) is Newton's Second Law which governs the motion of the pin. 

Geometric and constitutive relations used to close Eqs. (1-11) are as follows: 

V 1 = V, + V cpi +V gi , (12) 

(13) 
(14) 

(15) 

(16) 
(17) 

(18) 
(19) 



- 273 - 



v 2 = 


'■ Vcp 7 + Vg 2 i 


Z 


v 2 

p ~ A ' 


r[V s }-- 


/3V,\ 1/3 


A b [V 3 ] = 


(367TATV; 2 ) 173 , 


P 9i 


= PgiRT gi , 


n[P gi ] 


= -- = bP n 
dt 91 ' 


e,[T,]-- 


= f>/ei[T,], 



N cp 

€ cpi [T C pi] = /_j *cp e cp [Tcpi\ y 

i=i 

N g 

e gi [T 9i ] = Y,Y 9 3 ei[T gi ], 



i=i 



^[r 5 ] = ^y/^r(ei[r 5 ]), 

Nep d 

Cvcpi [T C pi] = 2^ <V~<JT \ e °P *■ C P*V ' 









i=i 



*c 



"-cpi L^cpiJ — 2-*t OP cp L^cpiJ i 



N 



^[^]^^[TJ, 



i=i 



Nc, 



c pcpi 1-^cpiJ = 2-f c p~JFt \ c p t c pi J / ' 

j-! ajt cpi 

Vcp,pi 1-^cpij^iiJ = hcp&Acpi {lepi ~ -tgi) i 

Qcpi = Qcpi [J-cpil ? 

ifP ff2 A p <F cr<t 



W ou t 2 — i^ 2 , , 



-i 






\ 



l(ft)""((ft)^-») »(%)<(*)* 



T£±I 



„(%)>(*>),-., 



JTcp 



1 -*7cp, 



m fl 



(20) 
(21) 
(22) 
(23) 
(24) 
(25) 
(26) 
(27) 

(28) 

(29) 
(30) 
(31) 

(32) 
(33) 

(34) 

(35) 



Here, and throughout the paper, braces [ ] are used to denote a functional dependence on 
the enclosed variable. Equations (12-14) are geometrical constraints; in Eq. (14), A p is 
the cross-sectional area of the pin. Equation (15) is an expression for the radius r of each 



- 27^ - 



spherical pyrotechnic grain; N is the total number of pyrotechnic grains. The area of the 
burn surface is given by Eq. (16); it is assumed here that the area of the burning surface is 
the total surface area of the N pyrotechnic grains. Equation (17) is a thermal equation of 
state for the gas phase products. Occurring in this expression are the gas phase pressure 
P 9i , the gas phase temperature T 5 -, and the ideal gas constant for the gas phase products R 
(the quotient of the universal gas constant and the mean molecular weight of the product 
gases). The pyrotechnic burn rate r& is given by Eq. (18). 

Equations (19-21) are caloric equations of state for the solid pyrotechnic, condensed 
phase products, and gas phase products, respectively. Here, T s is the temperature of the 
solid pyrotechnic, and T cpi is the temperature of the condensed phase products. Also, Y/, 
Yj p . , Yj. , N s , N cp , and N g are the constant mass fractions and number of component species 
of solid pyrotechnic, condensed phase product, and gas phase product species, respectively. 
Here, and throughout the paper, the notation superscript "j" is used to label quantities 
associated with individual chemical species. Since for both ideal gases and condensed phase 
species, the internal energy is only a function of temperature, the specific heat at constant 
volume for the solid pyrotechnic, c V5 , the condensed phase products, c vcpi} and the gas 
phase products, c vgi , can be obtained by differentiation of Eqs. (19-21) with respect to their 
temperature. Expressions for the specific heats at constant volume are given by Eqs. (22- 
24). The specific enthalpies for the condensed phase products and the gas phase products 
contained within the NSI assembly are given by Eqs. (25) and (26), respectively. These 
expressions can be differentiated with respect to their temperature to obtain the specific 
heats at constant pressure c pcpi and c P5l , Eqs. (27) and (28). 

Equation (29) gives an expression for the rate of heat transfer from the condensed 
phase products to the gas phase products. In this expression, h cPi3 is a constant heat 
transfer parameter, and A cpi is the surface area of the condensed phase products. The term 
h CPi9 A cpi is assumed large for this study. The functional dependencies of the heat transfer 
rates between the surroundings and the product subsystems are given by Eqs. (30) and 
(31). The functional form of these models will be given below. 

Equation (32) models pressure-volume work done by the gas contained within the ex- 
pansion chamber in moving the pin. Equation (33) models the force on the pin due to the 
gas phase pressure and a restraining force due to the shear pins which are used to initially 
hold the pin in place. Here, F cr i t is the critical force necessary to cause shear pin failure. 
The work associated with shearing the pin is not considered. 

The flow rate of gas phase product mass through the NSI port is given by Eq. (34). 8 
Occurring in this expression are the cross-sectional area of the port, A e , and the specific 
heat ratio for the product gases contained within the NSI assembly, 7 (= c P5l /c V5l ). This 
expression accounts for mass choking at elevated NSI assembly /gas expansion chamber 
pressure ratios. The condensed phase product mass flow rate through the port is given by 
Eq. (35). 

With the assumption of large heat transfer rates between the condensed phase and gas 
phase product subsystems (i.e., hcp^A^ — ► 00), the product subsystems remain in thermal 
equilibrium for all time. Therefore, we take T Pl = T cpi = T 9l and T P2 = T cp2 = T 52 , with T Pl 
defined as the temperature of the combined product subsystem contained within the NSI 
assembly and T P2 the temperature of the combined product subsystem contained within the 
gas expansion chamber. With this assumption, one can define the net heat transfer rates 
Q Vl and Q P2 governing the transfer of heat from the surroundings to the combined product 



- 275 - 



subsystems: 

Q Pl [T Pl ] = Qcp, + Q gi = hA Wl (T w - T Pl ) + *A Wl (aT* - eT^) , (36) 

Q n [TpJ = Q cp2 + Q g2 = hA W2 [V.] (T w - T P2 ) + aA W2 [V 2 ] (aT* - eT p 4 2 ) , (37) 

where 



^"^A/ 1 ' ""* " e ' ^ 2L " J "1A P 



Equations (38) are expressions for the surface area of the NSI assembly and the gas expan- 
sion chamber, respectively, through which heat transfer with the surroundings can occur; 
the parameter A\ in the first of these relations is the constant cross-sectional area of the 
NSI assembly. 

Mathematical Reductions 

In this section, intermediate operations are described that reduce the governing equa- 
tions to a final autonomous system of first order ODE's which can be numerically solved to 
predict the pin puller performance. To this end, it is necessary to define a new variable V 2 
representing the time derivative of the gas expansion chamber volume: 

*.**. (39) 

The final system consists of eight first order ODE's of the form 

£ = f(«). (40) 

where u = (V 2 ,V 8y V cpi , p gi ,T PlJ V cp2 ,T P2 ,V2) T is a vector of dependent variables and f is a 
non-linear vector function. These eight dependent variables will be referred to as primary 
variables. It will now be shown how to express all remaining variables as functions of the 
primary variables. 

Quantities already expressed in terms of the primary variables are the gas phase pressure 
inside the NSI P 9l [p gi ,T pi ], the heat transfer rates Q Pl [T Pl ] and Q P2 [V 2l T P2 ] y the specific 
internal energies e cpi [T Pi ] and e gi [T Pi ], the specific heats at constant volume c vcpi [T Pi ] and 
c vgi [T Pi ], the specific enthalpies /icpJTpi] and h 9l [T Pl ], and the specific heats at constant 
pressure c pcpi [T Pl ] and c pgi [T Pl ]. Also, with a knowledge of P gi , Eq. (18) can be used to 
express r& as functions of p gi and T Pl : 

rb[p gi ,T Pl ] = bP? 1 [p gi ,T Pl ]. (41) 

Addition of Eqs. (1), (2), (3), (7), and (8) results in a homogeneous differential equation 
expressing the conservation of the total system's mass: 

j f {psV, + PoVcn + p gi V gi + PcVcn + p gi V g2 ) = 0. (42) 

Integrating this expression, applying initial conditions, denoted by the subscript "o", using 
Eq. (12) to eliminate V gi in favor of Vi, V„ and V cpi , using Eq. (13) to eliminate V 92 in 
favor of V2 and V cp2 , and solving for p g2 results in the following: 

rT/ ,r T/ T/ , m - p a V, - p cp Vc Pl - p gi (Vi -V.-Vcpi)- PcpVcp 2 ( .„, 
Pg, [^2, V„ Vcpi , p gi , V CP2 \ V 2 -V ' ^ ' 

- 276 - 





dV 3 _ 
dt 


-Mn, 




dV cpi 


_ VcpPi 


*Mn - 


rh cp 


dt 




Pep 






dVcp 2 . 


_ rn C p 





where 

Wo = Ps^So "I" PcpVcpio + Pjio Vffio + PcpVcp 2 o + Pg2oVg20* 

Here, rn represents the initial mass of the system. Substituting Eq. (43) into Eq. (17) 
determines P 92 as a function of the primary variables: 

P 92 [V2> V„ V cpi , p gi , V CP2 , T P2 ] = p 92 [V 2 , K, V CP1 , p 5l , K P2 ] i?T P2 . (44) 

With a knowledge of P 32 , Eqs. (32), (33), (34), and (35) can be expressed in the following 
forms, respectively: 

w out2 = ^,V.,K w ,P ft ,V w r ft ,Vi], (45) 

F p = Fp^,^,^,^,^,^], (46) 

m g = mp^^^V^^p^^Tp^Fcp^TpJ, (47) 

"*cp = ™ cp [V2 , V*> Vc Pl » P51 ? Tpi > v cp 2 >T P2 \. (48) 

We next simplify the remaining mass evolution equations. Since both p 3 and p^ are 
constant, Eqs. (1), (2), and (7) can be rewritten as 

(49) 
(50) 

Hi ~ n ■ (51) 

at Pep 

To simplify Eq. (3), we use Eq. (12) to replace V gi in favor of V a and Vcpj, use Eqs. (49) 
and (50) to eliminate the resulting volume derivatives, and solve for the time derivative of 

d Pgi _ (*-» -(*-£) *»)/».*■- *»-&*■» 

The energy evolution equations will now be simplified. We first multiply Eq. (1) by e s 
and subtract the result from Eq. (4) to obtain 

de. 

-3f-°- 

Thus, in accordance with our assumption of no heat transfer to the solid pyrotechnic, its 
specific internal energy remains constant for all time. Integrating this result, we obtain 

e a = e so . (53) 

Addition of Eqs. (5) and (6), and addition of Eqs. (9) and (10) result in expressions governing 
the evolution of energy for the combined product subsystems contained within the NSI 
assembly and gas expansion chamber, respectively: 

^ [pcpVcp^cpx + PgxVgitgi] = Pt^aMn - h cpi m cp - h 9l m g + Q Pl , (54) 



- 277 - 



-j t [pcpV cp2 e cp2 + pg 2 V g2 € g2 ] = h cpi m cp + h gi m g + Q P2 - W out2 . (55) 

The net heat transfer rates given by Eqs. (36) and (37) have been incorporated into these 
expressions. Multiplying Eq. (2) by e cpi , multiplying Eq. (3) by e gi> subtracting these 
results from Eq. (54), using Eqs. (20) and (21) to re-express the derivatives in terms of T Pl , 
and solving for the derivative of T Pl yields: 

dT Pl _ p s (e s - r] cp e cpi - (1 - Tfe,) e gi ) A b r b - (/i cpi - e^ ) m cp - (h gi - e gi ) rh g + Q Pl ^ 

at pep ^cpi c vcpi + Pg\ Vg\ c vg\ 

(56) 
Similarly, multiplying Eq. (7) by e cp2 , multiplying Eq. (8) by e 52 , subtracting these results 
from Eq. (55), using Eqs. (20) and (21) to re-express the derivatives in terms of T P2 , and 
solving for the derivative of T P2 yields: 

dT P2 _ (h cpi - e cp2 )rh cp + (h gi - e S2 ) ra g + Q P2 - W oui2 ^ .^ 

dt PcpVcp2 € Cp2 C VCp2 » P92^92 C Vg2 

Lastly, Eq. (11) can be split into two first order ODE's. The first of these equations is 
given by the definition presented in Eq. (39). The second equation, obtained by using Eq. 
(39) and the geometrical relation given by Eq. (14), is expressed by the following: 

^ = Vi. (58) 

dt m p 

Equation (39), (49), (50), (52), (56), (51), (57), and (58) for a coupled set of eight 
non-linear first order ODE's in eight unknowns. Initial conditions for these equations are 

V 2 (t = 0) = V 2oy V s (t = 0) = V JO , V cpi (t = 0) = Vcp,*, 

p gi (t = o) = p, l0 , T Pl (t = 0) = T OJ V^t = 0) = V CP20 , (59) 

T P2 [t = 0) = T 0i V 2 (t = 0) = 0. 

All other quantities of interest can be obtained once these equations are solved. 

Results 

Numerical solutions were obtained for the simulated firing of an NSI into the pin puller 
device. The numerical algorithm used to perform the integrations was a stiff ODE solver 
given in the standard code LSODE. The combustion process predicted by the CET89 chem- 
ical equilibrium code followed the chemical equation given in Table 1. Parameters used in 
the simulations are given in Table 2. 

Predictions for the pressure history inside the NSI and the gas expansion chamber are 
shown in Fig. 3. Also shown in this figure are experimental values obtained by pressure 
transducers located inside the gas expansion chamber. 9 A rapid increase in pressure is 
predicted within the NSI assembly immediately following combustion initiation (t = 
ms); the pressure continually rises to a maximum value near 195 MPa occurring near the 
time of complete combustion (t = 0.023 ms). The pressure within the expansion chamber 
increases more slowly due to mass choking at the NSI port. Following completion of the 
combustion process, the pressure within the NSI assembly decreases to 53.9 MPa occurring 
near t = 0.06 ms; during this same time, the pressure within the gas expansion chamber 
uniformly increases to a maximum value of 53.4 MPa. There is a subsequent decrease in 

- 278 - 



both pressures" to values near 22.5 MPa at completion of the pin's stroke (t st = 0.466 ms). 
These decreases in pressure result from work done by the product gases in moving the pin 
and heat transfer from the combined product subsystems to the surroundings. 

Figure 4 shows the predicted temperature history for the combustion products contained 
within the NSI assembly and the gas expansion chamber. Since the only energy exchange 
between subsystems contained withih these two components is due to the flux of product 
mass through the NSI port, the resulting temperatures of the combined product subsystems 
do not thermally equilibrate. Figure 5 shows the predicted density history for the gas phase 
products inside the NSI and the gas expansion chamber. As a consequence of the product 
temperature difference, a significant difference in gas phase density is also predicted. 

The predicted velocity history of the combustion products flowing through the NSI port 
is given in Fig. 6. Here, a rapid rise in velocity to a maximum value near 928 m/s is predicted 
immediately following combustion initiation; during this time, the flow through the port 
becomes choked. The flow remains choked as the velocity slowly decreases to 830.3 rajs. 
Subsequently, there is a rapid decrease in velocity to a minimum value of approximately 
7 m/s ocurring at t = 0.63 ms. This rapid decrease in velocity occurs as the pressures 
within the NSI assembly and the gas expansion chamber equilibrate following completion 
of the combustion process. As the pin retracts, gases within the expansion chamber expand 
creating a slight pressure imbalance across the NSI port; consequently, the velocity of the 
flow begins to slowly increase to a value of 23 m/s at completion of the stroke. 

Figure 7 shows the time history of the predicted pin kinetic energy. A continual increase 
in kinetic energy to a maximum value of approximately 31.4 J at completion of the stroke 
is predicted. This value compares to an experimentally measured value of approximately 
22.6 J. The larger value for the predicted kinetic energy is consistent with the fact that 
frictional effects, which would tend to retard the motion of the pin, have not been accounted 
for in the model. 

Figure 8 gives results showing the sensitivity of the model to changes in the NSI port 
cross-sectional area, A e . For this study, we use the predicted pin puller solution as the 
baseline solution (baseline parameters given in Table 2). The sensitivity of the model is 
determined by solving the pin puller problem and finding the parametric dependency of 
three predicted quantities: the pin kinetic energy at completion of the stroke, the stroke 
time, and the maximum pressure attained within the NSI assembly. Quantities presented 
in this figure have been scaled by values obtained from the pin puller simulation presented 
above. For decreasing values of A e , pin kinetic energy decreases while both the stroke time 
and maximum pressure within the NSI increase. These results are primarily due to smaller 
mass flow rates through the port resulting from decreasing port cross-sectional areas. For 
slightly larger values of A e , both the stroke time and the pin kinetic energy approach a 
nearly constant value while the peak pressure within the NSI decreases. 

Conclusions 

The model presented in this paper is successful in predicting the dynamic events asso- 
ciated with the operation of an NSI driven pin puller. In addition to tracking the interac- 
tions between the reactant and product subsystems, the model also accounts for multiple 
pyrotechnic grains, variable burn surface area, and combustion product mass flow rates 
through the NSI port. Results of a sensitivity analysis reveal that variations in the cross- 
sectional area of the port may significantly effect the performance of the device. Specifically, 
significant decreases in the pin kinetic energy result from decreases in port cross-sectional 

- 279 - 



area. In the presence of friction, the Smaller kinetic energy of the pin may be insufficient to 
overcome frictional effects resulting in functional failure. Decreases in cross-sectional area 
may arise from the partial blockage of the NSI port by foreign matter or by the accumula- 
tion of condensed phase combustion products. Moreover, it is possible that the very high 
predicted pressures within the NSI assembly resulting from decreasing port cross-sectional 
areas may be sufficient to cause structural failure of the NSI's webbing, thereby jamming 
the pin and preventing it from retracting. Such structural failures have been reported in 
the past. 6 

References 

1 Razani, A., Shahinpoor, M., and Hingorani-Norenberg, S. L., "A Semi-Analytical Model 
for the Pressure-Time History of Granular Pyrotechnic Materials in a Closed System," 
Proceedings of the Fifteenth International Pyrotechnics Seminar, Chicago, IL, 1990, pp. 
799-813. 

2 Farren, R. E., Shortridge, R. G., and Webster, H. A., Ill, "Use of Chemical Equilibrium 
Calculations to Simulate the Combustion of Various Pyrotechnic Compositions," Proceed- 
ings of the Eleventh International Pyrotechnics Seminar, Vail, CO, 1986, pp. 13-40. 

3 Butler, P. B., Kang, J., and Krier, H., "Modeling of Pyrotechnic Combustion in an Automo- 
tive Airbag Inflator," Proceedings - Europyro 93, 5 e Congres International de Pyrotechnic 
du Groupe de Travaile de Pyrotechnic, Strasbourg, France, 1993, pp. 61-70. 

4 Kuo, J. H., and Goldstein, S., "Dynamic Analysis of NASA Standard Initiator Driven Pin 
Puller," AIAA 93-2066, June 1993. 

5 Gonthier, K. A., and Powers, J. M., "Formulation, Predictions, and Sensitivity Analysis of 
a Pyrotechnically Actuated Pin Puller Model," Journal of Propulsion and Power, accepted 
for publication, 1993. 

6 Bement, L. J., Multhaup, H. A., and Schimmel, M. L., "HALOE Gimbal Pyrotechnic Pin 
Puller Failure Investigation, Redesign, and Qualification," NASA Langley Research Center, 
Report, Hampton, VA, 1991. 

7 Gordon, S., and McBride, B. J., "Computer Program for Calculation of Complex Chem- 
ical Equilibrium Compositions, Rocket Performance, Incident and Reflected Shocks, and 
Chapman-Jouguet Detonations," NASA Lewis Research Center, SP-273, Cleveland, OH, 
1976. 

8 Fox, R. W., and McDonald, A. T., Fundamentals of Heat and Mass Transfer, 3 rd ed. John 
Wiley and Sons, Inc. 1985, pp. 599-617. 

9 Bement, L. J., private communication, NASA Langley Research Center, Hampton, VA, 
1992. 



- 280 - 



Table 1: Stoichiometric equation used in pin puller simulations. 



4.0908Zr(s) + 2.9178KCl0 4 (s) 



2J198Zr0 2 (cp) + 2A786KCl(g) + l.ZUOO(g) 
+1.2305ZrO 2 (g) + 1.0136O 2 (ff) + 0.6472C/(<7) 
+0.mOK(g) + 0.2434iirO(</) + O.UOZZrO(g) 
+0M88ClO(g) + 0.0285K 2 Cl 2 (g) + 0.0040tf 2 (<7) 
+0.003lCf a (y) + 0.0002Zr(g) + O.OOO10 3 (5)- 



Table 2: Parameters used in pin puller simulation. 



parameter 


value 


Vcp 


0.43 


N 


100 


A e 


0.100 cm 2 


A p 


0.634 cm 2 


Ax 


0.634 cm 2 


Vi 


0.125 cm 3 


Ps 


3.57 g/cm 3 


Pep 


5.89 g/cm 3 


T 3 


288.0 K 


T w 


288.0 K 


h 


1.25X10 6 g/s 3 /K 


e 


0.60 


a 


0.60 


Fcrit 


3.56 Xl0 7 dyne 


b 


0.003 (dyne/cm 2 )- - 60 ™/ s 


n 


0.60 


v 2o 


0.824 cm 3 


v, 


0.038 cm 3 


Vcp\o 


7.425 Xl0 -8 cm 3 


P910 


6.202 xl0~ 6 g/cm 3 


T 


288.0 K 


*CP2<> 


6.576X10" 7 cm 3 


v 2 


0.0 cm 3 js 



- 281 - 



200 I '' '''■■' ' I ' ' I I ' ' ' ' ' r i i i | i i i i i i i 



150 



^ 100 



50 



-0.10 



predicted result 



^— experimental result 




0.50 



Figure 3: Predicted and experimental pressure histories for the pin puller simulation. 



6000 



4000 



2000 



I i i i i | i i i i i i i i i | i ■ ' I 




i r ■ i I i 



-0.10 



0.00 



0.10 



0.20 

t(ms) 



0.30 



0,40 



0.50 



Figure 4: Predicted temperature histories for the pin puller simulation. 



- 282 - 



0.30 



r t t ? i i r r-r i i i i i i i i i i i i i t i i i i i i i i l r m r t ~ r-\ 




Figure 5: Predicted temperature histories for the pin puller simulation. 



i 



1000 


1 1 1 rV'l I 1 1 1 I I 1 T I "1 IT 1 1 I T ITTH IIIIIITItlfl 


T' ] "T I'T 1 ! I 1 1 1 1 1 I T I ! T T T T"T - 


800 


- 


^ 




- 


600 

s 

1 
1 


- 






- 


• 

r 

400 








~ 


200 








- 





.1 1 I..JL 1 1 1 1 1 









-0.10 0.00 0.10 0.20 0.30 0.40 0.50 

t(ms) 
Figure 6: Predicted velocity of the flow through the NSI port for the pin puller simulation. 



- 283 - 



^Q I I I I I I i i t I | I I i I i i i I I | I I I > t i i > i i i i I r t I I I I | i i i | t 



30 



tq 20 



10 



r< i i i i i i i i |_ 



I I 1 I I 1 1 I ■ I 1 I I > I I I I 1 I I I t I I 



-0.10 0.00 



0.10 



0.20 

t(ms) 



0.30 



0.40 



0.50 



Figure 7: Predicted kinetic energy of the pin for the pin puller simulation. 



2 
3 

8 



1 - 



- 


I 1 1 — 


—\ 1 1 1 — 1—| 1 1 1 1 1 T" "T 1 | 


^ 




4 


- 


Wt 


^^-^^^ \ 


- ^^ 


Jl 1 1 


t • > ...... i 



0.01 



0.10 



MK 



1.00 



Figure 8: Sensitivity of the model to changes in the NSI port cross-sectional area. The 
values presented in this figure have been scaled by the gaseline values A* = 0.10 cm 2 , 
KE* = 31.4 J, t* = 0.466 ms y and P^ = 195.2 MPa. 



- 28^ - 



FINITE ELEMENT ANALYSIS OF THE SPACE SHUTTLE 
2. 5 -INCH FRANGIBLE NUT 



J2V/5 



I 1 



/I 



Darin N. McKinnis 
NASA Lyndon B. Johnson Space Center, Houston, TX 



Abstract 
Finite element analysis of the Space 
Shuttle 2.5-inch frangible nut was 
conducted to improve understanding 
of the current design and proposed 
design changes to this explosively- 
actuated nut. The 2.5-inch frangible 
nut is used in two places to attach 
the aft end of the Space Shuttle 
Orbiter to the External Tank. Both 
2.5-inch frangible nuts must 
function to complete safe 
separation. The 2 . 5 -inch frangible 
nut contains two explosive boosters 
containing RDX explosive each 
capable of splitting the nut in 
half, on command from the Orbiter 
computers. To ensure separation, the 
boosters are designed to be 
redundant. The detonation of one 
booster is sufficient to split the 
nut in half. However, beginning in 
1987 some production lots of 2.5- 
inch frangible nuts have 
demonstrated an inability to 
separate using only a single 
booster. The cause of the failure 
has been attributed to differences 
in the material properties and 
response of the Inconel 718 from 
which the 2.5-inch frangible nut is 
manufactured. Subsequent tests have 
resulted in design modifications of 
the boosters and frangible nut. 
Model development and initial 
analysis was conducted by Sandia 
National Laboratories (SNL) under 
funding from NASA Lyndon B. Johnson 
Space Center (NASA-JSC) starting in 
1992 . Modeling codes previously 
developed by SNL were transferred to 
NASA-JSC for further analysis on 
this and other devices. An explosive 
bolt with NASA Standard Detonator 
(NSD) charge, a 3/4-inch frangible 
nut, and the Super*Zip linear 
separation system are being modeled 
by NASA-JSC. 

Introduction, 
The 2.5-inch frangible nut is used 
in two places to attach the aft end 
of the Space Shuttle Orbiter to the 
External Tank, as shown in figure 1. 
Each 2.5-inch frangible nut must 
function to complete safe separation 



of the Orbiter from the External 
Tank. Separation of each nut 
requires fracturing of four webs as 
shown in figure 2 . The 2 . 5-inch 
frangible nut contains two explosive 
boosters containing 100% RDX each 
capable of splitting the nut in 
half. To ensure separation, the 
boosters are designed to be 
redundant. The detonation of one 
-401 configuration booster, figure 
3 , is sufficient to split the nut in 
half. However, beginning in 1987 
some production lots of 2.5-inch 
frangible nuts demonstrated an 
inability to separate using only a 
single -401 booster. The cause of 
the failure has been attributed to 
differences in the material 
properties and response of the 
Inconel 718 from which the 2.5-inch 
frangible nut is manufactured. 
Details of the failure investigation 
were reported by Hoffman and 
Hohmann 1 . Subsequent tests have 
resulted in design modifications of 
the boosters and frangible nut. 

Finite element analysis of the Space 
Shuttle 2.5-inch frangible nut was 
conducted in cooperation with Sandia 
National Laboratories (SNL) , 
Albuquerque, New Mexico using two 
finite element analysis computer 
programs developed at SNL: JAC and 
PRONTO. JAC is a quasistatic finite 
element solver. JAC was used to 
simulate tensile pulls of Inconel 
718 in order to generate material 
characterizations of the Inconel 
718. The output of JAC can be used 
to qualitatively determine the 
advantage of one sample of Inconel 
718 over the other. JAC ' s output 
however can also be provided as 
input to PRONTO to improve the 
accuracy of booster and nut 
simulations. PRONTO is a dynamic, 
large deformation, finite element 
solver. PRONTO was used to conduct 
simulations of booster detonation 
and the response of the nut. These 
codes were primarily run at SNL, on 
a Cray Y-MP supercomputer. At NASA- 
JSC the codes were installed and run 
on a Sun workstation. 



- 285 - 



The first part of this paper 
discusses the material 
characterization process necessary 
for an accurate analysis. The second 
part of this paper discusses the 
structural analysis and design 
changes to the nut which were 
analyzed with comparison to test 
results. The third part of this 
paper briefly describes other 
devices which are being analyzed 
with this process. 

Inconel Material Characterization 
Process 
To conduct a structural analysis of 
the 2 . 5-frangible nut, material 
characterization of Inconel 718 is 
necessary. Material characterization 
requires a tensile test and a 
tensile test simulation using the 
JAC2D program. The procedure used 
is : 

1) Perform a tensile test. The 
tensile test must be carried out 
through failure, if at all possible, 
or at a minimum into necking. This 
is because reduction in area is a 
significant factor in the 
characterization process. 

2) Convert the tensile test 
engineering stress/strain data to 
true stress/strain over the range 
from yield to necking. Conversion of 
the data to true stress/strain is 
straight forward but necessary as 
the finite element programs use true 
stress/strain. True stress is found 
from engineering data according to 
the equation: 

true eng v eng 

True strain can be calculated 
according to the following equation. 



''true 



= ln(£ eng+ D 



Both equations are valid until 
necking begins, when the cross 
sectional area is no longer 
constant . 

3) Curve fit the true stress/strain 
data to a power- law hardening 
relationship, of the form 

C = G ys + A<£ p -e L >n 



to determine the hardening constant, 
A, and hardening exponent, n. O is 
the effective stress, C is the 
yield strength, £ p is the equivalent 
plastic strain, and £ L is the Luders 

strain. Inconel displays no Luders 
strain so this term can be 
considered zero. £ p should be 
evaluated according the Heavyside 
function, designated by the 
brackets, i.e. zero if its value is 
negative. 

4) Conduct a simulation of the 
tensile test using the JAC2D 
program. The finite element mesh for 
this simulation is shown in figure 
4. The left hand side is the initial 
mesh. The right hand side is the 
mesh near failure. Note that only 
the upper quadrant of the test is 
simulated due to two symmetry 
planes. A 2 mil reduction in 
diameter, which was measured from 
the test samples , and included in 
the mesh geometry assures 
localization at the symmetry plane 
on Z=0. 

5) Convert and review the results of 
the JAC2D analysis to obtain the 
tearing parameter. The results of 
the JAC2D analysis are converted in 
post-processing to the selected form 
of the tearing parameter, in this 
analysis it is 



TP 



£f 

"J: 



<2Gt) 



3(Gt-Om) 



-de 



where Ot 1S tne maximum principal 
stress, <J m is the mean stress, and 
£ is the equivalent plastic strain. 
This is evaluated from zero to £f r 
the strain at fracture. Knowledge of 
the final reduction in area from the 
tensile test is used here. The time 
step, tf f in the simulation is noted 
when the radius of the Z=0 symmetry 
plane equals the radius of the 
failed test sample. 

The tearing parameter can then be 
plotted versus time for the node at 
the center of the sample, where the 
tearing parameter is at a maximum in 
this model. The tearing parameter is 
then the value of the curve at tf. 
Using this method, tearing parameter 
sensitivity to time and/or reduction 
in area is displayed. 



- 286 - 



The yield stress and elastic modulus 
from the tensile test, and the 
calculated values of tearing 
parameter, hardening constant (A) 
and hardening exponent (n) are 
necessary to define Inconel 718 for 
the structural analysis to follow. 
The only other data required is the 
Poisson's ratio and density. These 
are all the parameters required by 
PRONTO' s constitutive model for this 
analysis. 

To confirm the results of the curve 
fit and JAC2D analysis, the JAC2D 
results can be plotted against the 
original tensile test data. If 
necessary, adjustments in A and n 
can then be made and another JAC2D 
analysis can be run until the 
analysis and tensile test agree. The 
tearing parameter can then be 
reevaluated. When the analysis and 
test agree, the material 
characterization process is complete 
and the structural analysis can be 
conducted on the 2.5-inch nut with 
the PRONTO program. 

Selection of the Tearing Parameter 
Fail nre Definition 
Chapter 16 of Numerically Modelling 
of Material Deformation Processes 2 
describes some of the prominent 
models that have been developed for 
ductile failure. However, in finite 
element analysis, most empirical 
models share the common approach of 
conducting an experiment or test on 
the material in question to 
determine the critical value, or 
tearing parameter, of the material. 
For example, the original model used 
a tensile test on a notched specimen 
to determine a tearing parameter. 
There are then different theories 
for which stress and strain 
components should contribute to the 
accumulation of the tearing 
parameter and in what empirical 
formulation. Apparently the correct 
formulation is strongly dependent on 
the material, geometry, and other 
factors for a specific problem, as 
well as the experience of the 
analyst . 

Using JAC, the calculation for 
tearing parameter is done in post- 
processing, providing complete 
flexibility to change the 
formulation of the tearing parameter 
calculation. The tensile test 
simulation does not even have to be 



rerun. New post processing commands 
are all that is required. PRONTO' s 
designers also anticipated the need 
for alternative constitutive models 
and tearing parameter formulations 
to meet the needs of the specific 
problem. And PRONTO supports post 
processing to permit reformulation 
of the tearing parameter if desired. 

However, NASA-JSC sought the ability 
to visualize in real-time the death, 
or failure, of material based on the 
tearing parameter. PRONTO supported 
adaptive or real-time death for 
energy, Vonmises stress, pressure, 
and other variables but not the 
tearing parameter. Therefore, the 
tearing parameter selected to 
describe failure of the Inconel 718 
in the 2.5-inch frangible nut was 
added to the elastic/plastic power 
law hardening material model of 
PRONTO specifically for this 
application. This new constitutive 
model, the power law hardening 
strength model, was based on the 
elastic/plastic power law hardening 
material model. The elastic/plastic 
power law hardening material model 
used by PRONTO was documented by 
Stone, Wellman, and Krieg^ . 

RDX Material Characterization 
Using PRONTO, explosives are modeled 
using Jones-Wilkins-Lee (JWL) 
parameters which were obtained from 
the Lawrence Livermore National 
Laboratories (LLNL) Explosives 
Handbook Properties of Chemical 
Explosives and Explosive Simulants 4 . 
The LLNL handbook also describes the 
formulation of the JWL parameters 
and their use to predict the 
"pressure-volume-energy behaviour of 
the detonation products of 
explosives in applications involving 
metal acceleration" . However, JWL 
parameters were not available from 
the handbook for 100% RDX, the 
explosive used by the booster. JWL 
parameters for 95% HMX were used 
instead. HMX is known to be slightly 
more energetic than RDX. 

structural Analysis - Phase I 
Nut and Booster Geometry - 
Initial structural analysis of the 
2.5-inch nut began in the spring of 
1992. At that time, tensile test 
data from yield through necking was 
not available for the Inconel 718. 
However, structural analysis was 
conducted to evaluate the 



- 287 - 



sensitivity of the nut to various 
geometrical factors. The finite 
element mesh currently being used is 
shown in figure 5. The slight 
differences between this mesh and 
the mesh used in phase I of the 
analysis are described later in this 
paper. Figure 6 shows in detail the 
side of the nut with a booster 
included where detonation will be 
modeled. The following changes in 
the nut geometry were analyzed to 
determine the effect on nut 
separation: radial gap between the 
booster cartridge material and the 
nut, outer notch depth of the nut, 
as shown in figure 7 , and the 
booster aspect ratio. The results 
were reported by K. E. Metzinger 5 . 

Radial gap between the booster and 
the nut is limited to 7 mil maximum 
by the tolerances on the nut and 
booster. The analysis and tests 
agree that minimizing the gap as 
much as possible, including the use 
of grease, is beneficial. Analysis 
shows the benefit is greatest in 
reducing gap from 7 mil to 4 mil, 5 
times greater than from 4 mil to 2 
mil. The advantage of reducing gap 
from 4 mil to 2 mil is the same as 
from 2 mil to . 5 mil. This suggests 
that tightening tolerances on the 
part to reduce gap probably would 
not be cost effective beyond 4 mil. 
Reduction of gap from 2 mil to . 5 
mil through the addition of grease, 
epoxy, or other agents would 
probably not justify the added 
complexity and cost of such a 
change . 

The outer notch depth of the nut has 
also been shown to be a factor in 
nut separation. The original flight 
configuration of the 2.5-inch 
frangible nut had an outer notch 
depth of 0.303 inches. Reducing the 
depth from 0.303 to 0.018 allows the 
nut halves to rotate further about 
the fourth web, from approximately 
25 degrees to over 60 degrees. 
Without a reduction, the corners of 
the nut at the outer notch pinch 
together, resulting in energy lost 
to compressive plastic strain. * 
Rotation of the halves places the 
elements of the fourth web, the last 
web to fail, under increasing 
tension. Increased tension means 
increased tearing values in those 
elements and improved likelihood of 
failure. This change was first 



demonstrated in test and repeated by 
the analysis. The outer notch depth 
was reduced to 0.075 in the -302 
configuration of the nut which is 
the current flight configuration. 

Although testing preceded analysis, 
the analysis showed two 
disadvantages to decreasing the 
outer notch depth: reduction in 
delivered energy to the nut from the 
booster and increased time until 
separation. Reducing the outer notch 
depth the nut causes web 1, the 
first web to fail, to fail earlier 
in time, before all the energy of 
the booster can be imparted to the 
nut. The analysis showed this loss 
to be small but significant if the 
outer notch depth is not properly 
sized. An outer notch depth of 0.153 
inches was actually shown to be less 
likely to separate than either a 
larger notch, 0.303 inches, or a 
smaller notch, 0.018 inches. The nut 
has not yet been designed to the 
optimum notch size. Further analysis 
will be required to determine the 
optimum size of the outer notch 
depth to ensure separation. 
Furthermore, the increased rotation 
permitted by the smaller notch depth 
causes the nut to fail later in 
time, in general. Although this is 
not a concern in the current design, 
this factor may be important in 
applications with smaller tolerances 
in separation time. 

Booster aspect ratio is defined as 
the diameter of the charge over the 
length of the charge. It was 
proposed that increasing the aspect 
ratio would increase the effect 
impulse delivered by the booster to 
the nut, with no additional RDX. 
Considerable energy was being lost 
or underutilized because it was 
located at or below the bottom of 
the webs of the nut and the 
separation was thought to be 
progressing in zipper fashion, from 
the top of the web toward the 
bottom. The advantage of increased 
aspect ratio was demonstrated in 
tests and analysis. As a result, 
NASA-JSC has modified the booster 
from the -401 configuration to the 
-402 configuration as shown in 
figure 3 for all future production 
of the booster. 



- 288 - 



Structural Analysis - Phase II 
A Copper Booster 
A second phase of analysis was 
conducted in September 1992 to study 
the effect of changing the booster 
cartridge material from stainless 
steel to copper. This study is 
briefly documented by K. E. 
Metzinger in a memo to D. S. 
Preece^. This study showed that a 
copper booster would absorb less 
energy in expanding within the 
booster port of the nut, making more 
energy available for nut separation. 
A nut would thus be more likely to 
fracture with a copper booster than 
with a stainless steel booster, as 
shown in figure 8 . 

The analysis also showed that the 
NASA Standard Detonator (NSD) , the 
initiating charge of the booster, is 
capable of blowing the top of the 
booster off and allowing the booster 
pressures to vent. This is confirmed 
by tests in which the booster top 
consistently separates from the 
assembly. Analysis showed that a 
copper booster would vent earlier 
than a stainless steel booster and 
the existing design does not permit 
strengthening in the area of 
fracture to prevent venting. 
However, venting is not considered 
to be a major concern as the 
analysis showed that the impulse 
delivered from the booster to the 
nut precedes the loss of the booster 
top. 

Unfortunately, copper has known 
compatibility problems with RDX. So 
a copper booster is not feasible. 
Testing has confirmed the analysis 
results by using modified stainless 
steel boosters, outer diameters 
machined down and copper sleeves 
inserted over the stainless steel, 
shown in figure 9. These boosters 
were successful in every test. 
Although these modified boosters 
have not been accepted for use in 
flight at this time, they have an 
advantage over a pure copper booster 
in that they would not increase the 
generation of shrapnel or increase 
venting because the top of the 
booster is unchanged from the flight 
boosters, solid stainless steel. 

Structural Analysis - Phase III 

pgfinina th e Model 

In 1993 the goal of the analysis was 

to improve the accuracy of the model 



so that design changes or 
variability between production lots 
of Inconel 718 could be compared 
quantitatively. Qualitative accuracy 
had already been demonstrated in the 
earlier analyses. To provide 
quantitative accuracy in the 
analysis it would be necessary to 
perform material characterization, 
as previously described, for 
different lots of Inconel 718 and 
correlate test performance with the 
analysis . 

The selected lots were HSX and HBT. 
It had been shown in earlier actual 
tests that these two lots of Inconel 
718 had significantly different 
characteristics based on their 
performance. No HBT nut has ever 
separated completely using the -401 
booster, while no HSX nut has ever 
failed to separate with a -401 
booster. The analysis had to show 
that HBT would not separate and HSX 
would not . 

The first step was to conduct a 
material characterization of HSX and 
HBT, then a structural analysis with 
the calculated material properties. 
The tensile test data for HSX and 
HBT is shown on figure 10. 

The model accurately predicted that 
HSX nuts would be much easier to 
fracture than HBT nuts. 
Significantly different tearing 
parameters, 0.345 for HSX and 0.675 
for HBT, indicated this trend. 
However, structural analysis 
conducted on HSX nuts failed to 
simulate a nut which completely 
separated. Clearly the model was not 
yet quantitatively accurate. Since 
the material characterization 
process with JAC analysis appeared 
sound, geometry and other 
characteristics were evaluated to 
increase the accuracy of the model 
and simulate a separating HSX nut. 
Three factors were found to increase 
accuracy of the model: the addition 
of a simulated bolt, increasing the 
granularity of the mesh in the webs, 
and increasing the number of points 
used to initiate detonation of the 
explosive. 

One of the refinements was that the 
number of elements in the webs was 
increased slightly. This is done by 
remeshing the nut's geometry. This 
reduces the size of the average 



289 - 



element in the web area of the nut. 
In PRONTO, the tearing parameter 
must be exceeded for the entire 
element for failure of that element 
to occur. A smaller element is thus 
more likely to fail due to localized 
increases in tearing parameter. 
Reducing average element size 
increases the number of elements 
which increases output file sizes 
and run times. Smaller elements also 
require smaller time steps, further 
slowing model processing. Thus this 
step, while increasing accuracy, 
increases run times, a common 
problem in finite element analysis. 

To increase accuracy of the model, 
the explosive material was provided 
with more initial detonation points. 
The original model had a single 
detonation point at time = 0. From 
this point the detonation wave was 
calculated to spread throughout the 
RDX. Ten detonation points were 
added to the model, each simulating 
detonation at time=0 . Thus the 
detonation of all the RDX occurs in 
a more even distribution and shorter 
time period. This produces slightly 
more output from the explosive and 
simulates a mature detonation wave 
as opposed to an initiating charge. 
This change did not require 
remeshing and was very easy to 
implement . 

A bolt was inserted into the hole of 
the nut. The addition of the bolt is 
significant to the separation of the 
nut, critical in some borderline 
cases, according to the analysis. 
Figure 11 shows the nut opening for 
2.5 mil radial gap and 4.5 mil 
radial gap HSX nuts. The 4.5 mil nut 
is opening similarly to the 2.5 mil 
nut until about 2 milliseconds. At 
that point opening has stopped and 
the nut is beginning to close down. 
However, at 3 . 5 milliseconds the nut 
which is moving to the left impacts 
the stationary bolt. The impact 
provided the necessary energy to 
complete the failure of web 4 and 
separation. At this time there is no 
experimental data to confirm this 
effect. However, testing is 
typically conducted with a bolt and 
no pre-load and should be included 
in the analysis. The effect of 
threads is not considered in the 
analysis. This change requires 
remeshing, the addition of a new 
material elastic Inconel 718, and 



the addition of contact surfaces. 
The number of elements added was 
kept to a minimum by treating the 
bolt as a perfectly elastic pipe. 

The simulation of an HSX nut with 
these changes resulted in a 
separating nut, failure of all four 
webs. Even with the changes which 
resulted in more energy to the nut 
and easier fracture, HBT nuts did 
not separate, still in agreement 
with tests. This study was 
documented in an SNL report^ . 

Future Studies of the 2.5-inch Nut 
A tearing parameter is used to 
determine when the Inconel 718 
material would fail. It has been 
proposed that the stainless steel of 
the booster should also be allowed 
to fail using a tearing parameter 
calculated by a JAC analysis. Even 
if the stainless steel is not 
permitted to fail, performing a JAC 
analysis, including a tensile test, 
should enhance the model's accuracy. 
Modeling conducted to date has not 
included material characterization 
of stainless steel as was done for 
the Inconel 718. NASA-JSC is 
currently conducting tensile testing 
of stainless steel samples from 
booster lots to perform this 
analysis . 

One geometrical design consideration 
which has not been modeled is the 
application of a backing plate or 
washer to the nut. During testing 
this was shown to have significant 
effect on the separation of the nut, 
overshadowing most other variables. 
Nuts without the backing plate, were 
not as likely to separate. However, 
to include the backing plate the 
model must be converted into three 
dimensions. NASA-JSC plans to 
conduct these analyses in 1994. 
NASA-JSC will probably run this 
analysis on their Cray to handle the 
significant increase in the number 
of elements necessary to conduct 
this analysis. 

Structural Analysis and Material 
Characterisation Conclusions 
The most significant conclusions 
from the analysis performed on the 
2.5-inch frangible nuts are as 
follows: 

A. Radial gap between the booster 
and the frangible nut is an 



- 290 - 



important dimension. This gap should 
be minimized if complete separation 
is desired. 

B. The energy absorbed by the 
stainless steel booster is 
significant. Switching from an all 
stainless steel cartridge to a 
cartridge of reduced diameter and 
the addition of a copper sleeve 
should increase the impulse 
delivered to a nut. An all copper 
booster housing is not recommended 
due to compatibility issues between 
copper and RDX. However, a hybrid 
booster made of stainless steel with 
a copper sleeve has been shown in 
test to be effective. 

C. Reduction in area obtained from a 
tensile test of an Inconel 718 
sample is significant . Reduction in 
area is a significant factor of the 
tearing parameter. Inconel with a 
relatively small reduction in area 
will be easier to break. 

Other Applications 
NASA-JSC is currently pursuing 
analysis models of the Space Shuttle 
3/4-inch frangible nut, the 
Super*Zip linear separation system, 
and a JSC-designed explosive bolt 
utilizing the NASA Standard 
Detonator (NSD) as the actuating 
charge. These models can use the 
same process as was used for the 
2.5-inch frangible nut with slight 
variations. 

The mesh for the explosive bolt 
model is shown in figure 12. Six 
prototype explosive bolts have been 
fired to date. Agreement with the 
analysis has been excellent. The 
analysis has provided design 
modifications which will be used to 
slightly change the separation plane 
and ensure positive retention of the 
bolt in its confining washer. 

The mesh for the Super* Zip linear 
separation system model is shown in 
figure 13. 

Conclusions 
The test and finite element analysis 
methodology developed by SNL and 
NASA-JSC has been successfully 
demonstrated. This methodology 
requires the use of SNL developed 
software which has been successfully 
transferred, with substantial 
assistance from SNL, to NASA-JSC. 



Training in the use of these codes 
has been provided. This methodology 
can now be used by NASA-JSC in 
several ways . 

Analysis can be conducted on new 
production lots of Inconel 718. 
Inconel 718 lots which possess 
material properties and tearing 
parameters outside of acceptable 
limits can be rejected before 
expensive machining and acceptance 
testing is conducted. 

Analysis can be conducted on sample 
lots of Inconel 718 to determine the 
effect of various heat treatment 
procedures on the microstructure. 
This approach will suggest measures 
to further control the forging 
process . 

Analysis can be conducted to 
quantify the effects of proposed 
design changes before manufacturing 
and testing is initiated. Analysis 
could also be used to establish the 
design margin of the current 
configuration or select a more 
meaningful design margin criteria. 
Currently design margin is 
determined by increasing the web 
thickness to 120% of the maximum 
allowable, as shown in figure 14. 
This choice for margin demonstration 
is not ideal. It is unlikely that 
the web thickness will be 
incorrectly manufactured and that 
this error will be overlooked in 
acceptance inspections . Furthermore , 
no additional information beyond 
separation versus failure to 
separate is obtained. It has been 
suggested that velocity of the nut 
halves as the nut separates would be 
a better demonstration of margin 
where failure to separate provides a 
velocity of zero. At this point, 
test methods including breakwires 
and high-speed photography are being 
used to determine the velocity of 
nut halves for comparison with the 
analysis. 

The test and analysis methodology is 
general enough that it can be 
successfully applied to other 
mechanical devices and design 
problems. NASA-JSC is currently 
pursuing analysis models of the 3/4- 
inch frangible nut, the Super*Zip 
linear separation system, and a 
NASA-JSC designed explosive bolt 



- 291 - 



utilizing the NSD as the actuating 
charge . 

References 

1. Investigation of Failure to 
Separate an Inconel 718 
Frangible Nut, William C. 
Hoffman, III, and Carl Hohmann, 
NASA Lyndon B. Johnson Space 
Center , Houston , TX . 

2. Numerical Modelling of Material 
Deformation Processes, Edited by- 
Peter Hartley, Ian Pillinger, 
and Clive Sturgess, Springer- 
Verlag, Germany, 1992 . 

3. A Vectorized Elastic/Plastic 
Power Law Hardening Material 
Model Including Luders Strain, 
Charles M. Stone, Gerald W. 
Wellman, Raymond D. Krieg, 
Sandia National Laboratories, 
Albuquerque, New Mexico, SAND90- 
0153, March 1990. 

4. LLNL Explosives Handbook 
Properties of Chemical 
Explosives and Explosive 
Simulants, B. M. Dobrantz and P. 
C. Crawford, UCRL-52997. 
Lawrence Livermore National 
Laboratory, Livermore, 
California, January 1985. 

5. NASA Frangible Nut Preliminary 
Findings, Memo from K. E. 
Metzinger to D.S. Preece, Sandia 
National Laboratories, 
Albuquerque, New Mexico, August 
25, 1992. 

6. NASA Booster Cartridge Material, 
Memo from K. E. Metzinger to 
D.S. Preece, Sandia National 
Laboratories, Albuquerque, New 
Mexico, October 5, 1992. 

7. Structural Analysis of a 
Frangible Nut Used on the NASA 
Space Shuttle, Kurt E. 
Metzinger, Sandia National 
Laboratories, SAND93-1720, 
November 1993 . 



modeling. The author also wishes to 
thank Carl Hohmann, NASA-JSC, who 
recommended many of the design 
modifications and variables which 
were studied and supported the 
interpretation of the analysis. 



Acknowledgments 
The author wishes to acknowledge the 
work of D.S. Preece and Kurt E. 
Metzinger, Sandia National 
Laboratories, Albuquerque, New 
Mexico, in model generation, initial 
modeling and analysis, and in 
technical support for NASA-JSC 



- 292 - 




ORBITEFVEXTERNAL TANK 
AFT ATTACH INTERFACE 




Enforced /-direction 
Displacement 



S 

p //\\ 

•j "TrtTi 

si ~t T i n ' 

o IJlj j lflijlJ Qf 

z H l UI-l iit i ll t l 



No /-direction 
Displacement 



--'0 -0 -10 .20 -.10 .00 .10 .20 



"jjTfYP 


IIHIIIUMIIMfl 
ItltMlllldllfifl 





Figure 1. Location of the 2.5-inch 
frangible nut between the Space Shuttle 
Orbiter and External Tank. 



Figure 4. Tensile test mesh for 
simulation by JAC2D. 




TOP VIEW OF 2.5-INCH 
FRANGIBLE NUT 



FRANGIBLE WEBS 




SECTION A-A 
FRANGIBLE NUT 
SEPARATION PLANE 




-3.0 2.0 i. a 



1.0 2. a 3.0 



Figure 2 . Top view and side view of the 
2.5-inch frangible nut. 



ISOMICA 
DISKS 




1920 mg , 

RDX — I 



HOUSING 




1920 mg 
RDX 



HOUSING 



-401 CONFIGURATION 



-402 CONFIGURATION 



Figure 3. Side view of the 
2.5-inch frangible nutbooster, 
-401 and -402 configurations. 



Figure 5. 2.5-inch frangible nut mesh. 



Y'-!v'T.ui;f : T : f..V;;:.:o - • .' 




t* ttt t 

Poinl A | 



Web 1 

*— £jq»losive 
Cartridge 



_ _„.1_. .1 L. .-..A J. J 

00 1.25 ISO 1.75 2.00 2.25 2. SO 2.75 3.00 



Figure 6. 2.5-inch frangible nut mesh, 
detailed view. 



- 293 - 




TOP VIEW OF FRANGIBLE NOT 




SECTION A 



Figure 7. Outer notch depth of the nut. 



0.60 


T "" i 


— ■ 1 .- T ~ 


0.50 


S^" 


0.40 




Copper 

304L 


a 

£ 0.30 

9 




■ 


1 






0.20 




■ 


0.10 


/>'" 


i _i 



0.000 



0.002 0.003 

Time (sec) 



0.00£ 



Figure 8 . Nut opening as a function of 
booster material, copper versus 
stainless steel. 



~Uh\ 







J qr_rvw\ 



1 Copper sleeve 

Figure 9 . Stainless steel booster and 
stainless steel booster with copper 
sleeve. 



250x10 -I 



200- 






150- 










CO 


• 


? 


• 


■c 




CD 
01 


100- 


c 


• 


? 


• 


LU 


■ 



50- 



1 



0.00 



Typical Tensile Test Data for Inconel 718 Lots HBT and HSX 
(.005/rnin strain rate) 



^HSX 
— HBT 



i I i i i i | i i 

0.05 0.10 

Engineering Strain (inches/inch) 



1 I ' ' 

0.15 



i j i i i i | i — i i i | 

0.20 0.25 0.30 



Figure 10. Tensile test data for Inconel 718 lots HBT and HSX. 



- 294 - 



I 




2.0 30 
Time (ms) 

Figure 11. Nut opening as a function of 
gap. 



Explosive bolt body 




Confining washers 



ration plane 



explosive charge 



ZZZZZZZZ3\ ai . 25 " 



zzzzzzza : 



T 



-301 FRANGIBLE NUT WEBS 



ZZZZZZZ3 



115% (0.1 49") 
120% (0.156") 



Z22ZZZZ3I 



-1 01 , -1 02 MARGIN NUT WEBS 

Figure 14. Web dimensions for 
2.5-inch frangible nuts, -301 
flight configuration, and -101 and 
-102 margin configurations. 



Figure 12 . Finite element mesh for the 
explosive bolt with NSD charge. 




Aluminum doublers 

Lead sheath confinement 

HMX explosive cords, end on 



Figure 13. Super*Zip linear separation 
system finite element mesh. 



- 295 - 



ANALYSIS OF A SIMPLIFIED FRANGIBLE JOINT SYSTEM 



514- IS 

~Joo4- 



Steven L. Renfro 

The Ensign-Bickford Company 

Simsbury, CT 

James E. Fritz 

The Ensign-Bickford Company 

Simsbury, CT 



Abstract 

A frangible joint for clean spacecraft, fairing, and stage 
separation has been developed, qualified and flown 
successfully. This unique system uses a one piece 
aluminum extrusion driven by an expanding stainless steel 
tube. A simple parametric model of this system is desired to 
efficiently make design modifications required for possible 
future applications. Margin of joint severance, debris control 
of the system, and correlation of the model have been 
successfully demonstrated. 

To enhance the understanding of the function of the joint, a 
dynamic model has been developed. This model uses a 
controlled burn rate equation to produce a gas pressure wave 
in order to drive a finite element structural model. The 
relationship of the core load of HNS-IIA MDF as well as 
structural characteristics of the joint are demonstrated 
analytically. The data produced by the unique modeling 
combination is compared to margin testing data acquired 
during the development and qualification of the joint for the 
Pegasus* vehicle. 

Introduction 

Frangible joints have been demonstrated 
as robust and contamination free 
separation systems for various spacecraft 
and launch vehicle stage and fairing 
separation. Typical frangible joint 
systems are initiated using mild 
detonating fuse (MDF) detonation 
products to expand an elastomeric 
bladder which then compresses 
dynamically against a formed stainless 
steel tube. The high pressure developed 
at the tube forces it to a more round 
shape in order to fracture an aluminum 



plate along a stress concentration 
groove. This fracture provides 

separation without fragmentation or 
contamination because the products are 
contained within the steel tube. A typical 
joint cross section is depicted in Figure 1. 

Integrating this technology into new 
systems, with more challenging 
environmental conditions, could benefit 
from analytical modeling to properly 
configure each system. Understanding 
the mechanism required to sever the 
aluminum extrusion is crucial to meet 
new system requirements with full 
confidence. 

The purpose of this report is to 
document The Ensign-Bickford 
Company's efforts to develop a simple 
analytical tool using widely available 
hydrodynamic and finite element 
computer codes. 

Background 

The ANSYS 5.0 # finite element software 
allows for transient input to structures in 
the frequencies expected during a small 
damped detonation event. The frangible 
joint geometry is believed suitable for this 
type of analysis. To generate transient 
pulses for input into the finite element 



® Pegasus is a Registered Trademark of Orbital Sciences Corporation 
• ANSYS is a Registered Trademark of Swanson Anaysis Systems, Inc. 



- 297 - 



J-fy 



model, simple one-dimensional 
hydrodynamic analysis is used. 

For a one-dimensional Lagrangian model, 
a cylindrical geometry was assumed. 
The SIN 1 hydrodynamic code was used 
to solve conservation equations of 
momentum, mass, and energy. In 
order to use this information as input for 
the finite element model, individual or 
groups of cells were monitored to 
develop input equations for the finite 
element calculations. 

Since the hydrodynamic analysis is one 
dimensional, it limits the amount of 
understanding developed regarding the 
specific stress state existing in the 
aluminum. Peak stress locations and 
probable points of secondary failure 
cannot be determined, and assumptions 
must be made for the stiffness and 
response characteristics. A two 
dimensional hydrodynamic analysis 
would assist in understanding these 
effects, but such analysis is time 
consuming, and requires access to 
sufficient hardware and software 
resources. Also, the hydrodynamic 
model is unable to account for a wide 
range of thermal loads and structural 
preloads in the parts, and cannot be 
used to evaluate stresses in the part due 
to events other than the explosive 
loading (i.e., flight loads, thermal 
response, assembly loading). 

To understand the dynamic response of 
the frangible joint, an ANSYS 5.0 finite 
element model was created for use in a 
non-linear transient analysis. This 
analysis was used to determine the 
dynamic response of the aluminum when 
subject to transient loads driving the 
material above its yield strength. This 
methodology had been successfully used 



by The Ensign-Bickford Company to 
solve problems involving explosively and 
pryotechnically loaded structures. This 
paper represents the first time this 
technique used input developed by a one 
dimensional hydrodynamic analysis code. 

Model Development 

Using the SIN analysis, the critical areas 
were determined for input into the 
ANSYS 5.0 model. The timing and 
reaction of the shock waves incident and 
reflected from the interior steel wall result 
in two distinct types of relationships, both 
of which are decaying sinusoidal 
functions. 

The area of the stress riser has extremely 
high initial amplitude which rapidly 
decays. This is consistant with the 
geometry present at this location. A thin 
layer of elastomeric material and a thin 
aluminum section bounding the steel do 
not support reflected pressure waves as 
well as the thicker off axis areas. 
Basically, the initial detonation front 
experiences a rapid ring down within the 
wall of the steel tube. The inside surface 
of the steel responds approximately as 
illustrated in Figure 2. 

The second input function used is from a 
cross section at 45° from the stress 
riser. This relationship was similar in 
frequency to Figure 2 with a much lower 
initial amplitude. Figure 3 shows this 
relationship. 

A logical choice for a third function is 90 ° 
to the separation plane. Most frangible 
joint designs use air gaps combined with 
thin silastic sections to control position of 
the MDF and to allow easy installation. If 
no air gaps are assumed, the resulting 
function resembles the data in Figure 3 



- 298 - 



with lower amplitude. Introducing the air 
gaps increases the difficulty of the 
hydrodynamic analysis without any real 
benefit. This third function was therefore 
not used for this simplified approach. 

The source explosive used for this 
particular design is HNS-IIA. The 
equation of state for HNS is not currently 
available as part of the SIN database. 
Alternate explosive materials were used 
to bracket the response of HNS. The 
density, chemistry, detonation velocity, 
and Chapman-Jouget pressure were 
matched as closely as possible with 
candidate materials from the SIN 
database. Table 1 lists the explosives 
used and their properties compared to 
HNS. 

To simplify the ANSYS model geometry, 
symmetric constraints are used along the 
notch edge and one half of the joint 
mounting flange. The length of the 
flange was shortened to reduce the 
number of degrees of freedom which 
needed to be incorporated into the 
model. This model is shown in Figure 4. 

The transient loads are applied as 
pressure pulses along the interior of the 
aluminum. These loads have the same 
time profile as that predicted by the one 
dimensional hydrodynamic analysis, 
however input pressure amplitudes are 
reduced to achieve numerical stability. 
Unfortunately, this assumption is 
required, although it is expected that the 
results still allow development of an 
understanding of the aluminum response. 
The aluminum material (6061 -T6) was 
assumed to act in an elastic-perfectly 
plastic manner. That is, once the yield 
strength of the material is exceeded, no 
additional load can be supported by that 
material. Plastic convergence is 



achieved using the Modified Newton- 
Raphson method, based on a Von-Mises 
yield criterion. For the transient portion 
of the analysis, the Newmark time 
integration scheme is utilized, using 
Rayleigh damping with only mass matrix 
contributions (Beta damping). This 
applied damping is necessary in order to 
provide stability of the solution. 
However, a Beta term is chosen which 
ensures a low level of damping (0.05% or 
less) above 10,000 Hz. 

For the initial time steps, the symmetry 
constraints are applied to both the top 
notch and the flange edges of the model. 
When sufficient stress levels are 
determined in the notch to induce section 
failure, this symmetry constraint on the 
notch is removed and the leg of the 
section is allowed to bend up and away 
from its initial position. This is done to 
simulate proper function of the joint 
during the explosive event. 

Results of Finite Element Model 

As part of the preliminary work 
performed using this model, simple static 
stress analysis (linear and non-linear) as 
well as modal analysis were performed to 
verify model integrity and to learn about 
the basic structural characteristics of the 
model. Some important data was 
gleaned from these runs, including the 
presence of a potential plastic hinge near 
the flange region of the aluminum 
structure. Additionally, the modal runs 
showed that the aluminum had its 
second, third, and fourth normal modes 
between 50 and 200 kHz. This was 
important information, since it showed 
that the aluminum is capable of dynamic 
elastic structural response near the input 
frequency of the shock pulse. The 2 nd , 
3"*, and 4 th , mode shapes are shown in 



- 299 



Figure 5. 

Once the simpler analysis had been run 
and verified with hand calculations, the 
more complex non-linear transient 
analysis was run. The input pulse was 
characterized as a shock pulse with a 1 .5 
/xsec rise and a 1.5 /xsec decay at the 
notch location. The magnitude of this 
pressure pulse was chosen to remain 
slightly below the yield point of the 
material (approximately 36 ksi) to avoid 
model stability problems. As noted 
above, this assumption needed to be 
made, however; much information about 
the dynamic response of the structure 
was still learned. 

The final results are illustrated in Figure 
6. During the rise time of the initial 
pressure pulse, the structure cannot 
significantly respond to the high 
frequency input. The structure simply 
transmits the shock wave through the 
material thickness. By the time the pulse 
is damped, the structure begins to 
significantly respond, and peak stresses 
in the notch exceed the allowable 
material strength. A plastic condition 
through the wall is reached. It is at this 
time that separation occurs, and the 
symmetric boundary condition along the 
notch edge is removed. After this time, 
the load is no longer applied and the 
inertial loads of the aluminum leg are all 
that is left driving the deflection. 
Obviously, the loading assumption is 
somewhat non-realistic; the shock wave 
applied to the aluminum will continue 
along the inner wall even after separation 
has occurred. 

After this, the frangible joint is behaving 
as a cantilever beam with a fixed edge 
along the mounting flange. A plastic 
zone develops along much of the length 



of the flange wall. It is interesting to note 
that a plastic hinge develops in the bend 
region of the aluminum leg. This hinge 
location corresponds well to explosive 
over tests where a section of the 
aluminum became a flyer. 

Finally, at 24 /xsec, the leg has plastically 
deformed over its entire length, a plastic 
hinge occurs near the top of the 
extrusion, and model convergence is no 
longer possible using the elastic-perfectly 
plastic static strength allowable. The 
predicted deflection at this time is 0.103 
inches. For the actual hardware, it would 
be expected that energy would be 
expended by bending at the plastic hinge 
until the impulse had been dissipated. 

Discussion 

The aluminum is capable of responding 
to the input shock pulse in the 2 to 3 
Msec regime, suggested by the modal 
analysis and supported by the transient 
analysis. At approximately 3 /xsec after 
the shock pulse has arrived at the interior 
of the aluminum stress riser, failure at the 
groove is expected to occur. There 
remains sufficient energy to severely 
deform the legs once the failure at the 
stress riser has occurred. A secondary 
plastic hinge forms at the bend joint near 
the mounting flange for this particular 
design. 

The aluminum cross section is very 
efficiently dissipating the applied impulse 
once the stress riser failure occurs. In 
other words, plastic stresses do not 
localize and exist over much of the inner 
and outer surfaces of the aluminum. 

All of these discussion items show good 
agreement with test specimen articles. 
No failures of this particular joint have 



- 300 - 



occurred which would disagree with the 
conclusions of this analysis. 

The hydrodynamic analysis would 
provide much better resolution if a two 
dimensional model were used. Digital 
resolution of individual cell results could 
be used as forcing function for the finite 
element techniques. 

Although not specifically addressed by 
this paper, the one dimensional 
hydrodynamic analysis combined with 
the two dimensional ANSYS analysis 
shows good promise for evaluation of the 
effects of thermally induced strains and 
launch load induced stresses. A two 
dimensional hydrodynamic input would 
further enhance the ability of this 
technique to simulate flight functional 
conditions. 



References: 
1) 



2) 



Charles L Mader: Numerical Modeling of 
Detonations : University of California 
Press; 1979; pp. 310 -332. 

B.M. Dobratz; LLNL Explosives 
Handbook; UCRL-52997, 



301 - 




Figure 1 Frangible Joint Before and After Function 



- 302 - 




-I r 

5 10 

Time From Detonation of MDF (usee) 



Figure 2 Pressure Time History at Interior of Steel Tube at Stress Riser 



- 303 - 



0.02' 



to 

-Q 



~ 0.015- 



o 
2 



- 0.01- 



4> 

8 



« 0.005- 



u 
o 

.c 
CO 



0- 



4> 

i -0.005-I 

Q. 

E 
< 



-0.01 




10 20 30 

Time From Detonation of MDF (usee) 



40 



Figure 3 Pressure Time History at 45" From Stress Riser 



- 304 - 









m 



P; 






****?& 



1*: 



. If ! 






Y 

Ml 



k_nK 



NX 



Figure 4 ANSYS Finite Element Model 



- 305 - 



MODE 2 
52.827 HZ 




MODE3 
102.348 HZ 



MODE 4 
154.759 HZ 



Figure 5 ANSYS Finite Element Model Elastic Normal Modes 



- 306 - 




FEB 7, 94 

6:08:03 
NODAL SOLUTION 
STEP«3 
SUB «56 
TIME=.140E-04 
SEQV ( AVG) 

DHX «. 103667 
SUN =606.504 
SKX =37224 
A -0 
B =3660 
C =7300 
D =10950 
E =14600 
F =18250 
6 =21900 
H =25550 



=32650 
=36500 



*.':**»:*k&^''>: 



m . : - . :■. " i " "*"* • '" " ^J ' ' • ' -• ' 1 ! ^3 • ' -- ' ■ ' ■"* ' - ' • ' ■' 1 - l -£ ■ ■ ■ ■xz^2Qs<^ ~ Lx/ - ; ' Jk ' ^ii»L-^iJiA ^ 



Figure 6 Results of Hydrodynamic and FEA Combined Model 



- 307 - 



Table 1. Explosive Properties Used to Bracket HNS Performance 2 



Material 


Chemical 
Formula 


Density 
(9/cm 3 ) 


Detonation 

Pressure 

(kbar) 


Detonation 

Velocity 

(mm/Msec) 


HNS 


C 14 H 8 N 6 12 


1.60 


200 


6.80 


TATB 


C e H 6 N 6 6 


1.88 


291 


7.76 


TNT 


C 7 H 5 N 3 O e 


1.63 


210 


6.93 



- 308 - 



UNLIMITED DISTRIBUTION 

PORTABLE, SOLID STATE, FIBER OPTIC COUPLED 
DOPPLER INTERFEROMETER SYSTEM FOR DETONATION AND SHOCK 

DIAGNOSTICS 

K. J. Fleming, O.B. Crump 

Sandia National Laboratories 

Albuquerque, New Mexico, 87123 






i 



VISAR (Velocity Interferometer System for Any Reflector) is a specialized Doppler 
interferometer system that is gaining world-wide acceptance as the standard for 
shock phenomena analysis. The VISAR's large power and cooling requirements, and 
the sensitive and complex nature of the interferometer cavity have restricted the 
traditional system to the laboratory. This paper describes the new portable VISAR, 
its peripheral sensors, and the role it played in optically measuring ground shock of 
an underground nuclear detonation. The Solid State VISAR uses a prototype diode 
pumped Nd:YAG laser and solid state detectors that provide a suitcase-size system 
with low power requirements. A special window and sensors were developed for 
fiber optic coupling (1 kilometer long) to the VISAR. The system has proven itself 
as a reliable, easy to use instrument that is capable of field test use and rapid data 
reduction using only a notebook personal computer (PC). 



INTRODUCTION 

Detailed analysis and accurate models of shock 
phenomena and high speed motion require an 
instrument that is capable of measuring the high 
acceleration of surfaces accurately and non- 
intrusively. Dent blocks and stress gauges can only 
infer the final velocity of detonations while critical 
information pertaining to the acceleration is 
unknown. A versatile instrument that optically 
measures acceleration, displacement and velocity is 
VISAR. VISAR (Velocity Interferometer System 
for Any Reflector) uses coherent, single frequency 
laser light to illuminate a target that has some 
reflectivity. The reflected light is collected and 
routed through a modified, unequal leg, Michelson 
interferometer. As the target moves, the resulting 
Doppler information is detected and electronically 
analyzed, then the data are converted to velocity and 
displacement time histories using software operating 



on a personal computer (PC). The sensitivity, 
accuracy, and high bandwidth of VISAR is 
attributed to the optical method of measurement and 
its 400 MHz bandwidth is primarily limited only by 
the electronics in the system. 

Although the VISAR technique is excellent for 
measuring shock phenomena, there are some 
limitations with the conventional VISAR, such as; 
inherent sensitivity to adverse environments found 
outside the laboratory, hazardous unenclosed laser 
beam, high current, voltage and cooling 
requirements, and an inability to measure devices not 
in the "line of sight" of the laser beam, e.g. through 
smoke, tunnels and inside chambers. In an attempt 
to improve on the versatility of VISAR, a solid state 
system with fiber optic coupling and rugged 
components has been developed and rigorously 
tested in harsh environments. The fiber optic 
coupled sensor used to send and collect the light at 



- 309 - 



the target is unique from previous techniques and, in 
recent tests, has performed flawlessly even after four 
months encapsulation in curing concrete. The solid 
state VISAR described in this paper was used to 
measure ground shock generated by a nuclear 
detonation at the Nevada Test Site (NTS). 

BACKGROUND 

The VISAR was developed by Barker and 
Hollenbach 1 primarily for measuring free surface 
velocities of materials in gas gun experiments. An 
improved version of VISAR, developed by 
Hemsing^, electronically inverts and adds the 180° 
out-of-phase optical signals that were previously 
wasted, which effectively cancels target self-light 
and doubles the signal intensity. During fast shock 
jumps, the system may miss Doppler information 
which can cause discrepancies in measured velocity. 
For this reason, Kennedy and Crump 3 developed the 
double-delay-leg system. The double-delay- VISAR 
takes the return light, splits the optical signal and 
routes it through two interferometer cavities with 
different sensitivities. The data can then be more 
accurately reduced by comparing the results of the 
two systems. The conventional VISAR has many 
sensitive components mounted on an optical table. 
The Fixed-Cavity VISAR, developed by Stanton, 
Crump and Sweatt 4 , simplifies the interferometer 
cavity by cementing the movable components 
together. The result is a rugged, small, easy to use 
system with a minimal amount of adjustment. 

A low cost, portable VISAR using a diode laser and 
a fiber optic coupled sensor has been developed by 
Fleming, and Crump 5 with successful velocity 
measurements taken on electrically initiated slappers. 
The diode's invisible laser light makes alignment to 
the target difficult and the aberrated, high 
divergence beam profile of the laser is difficult to 
propagate through space for any extended length. 
Both problems are solved by the development of an 
imaging fiber optic coupled sensor^ that has intra- 
optic video capabilities. The sensor allows for 



remote target measurements and verification of the 
correct area of target illumination, (figure 1). 



IMAGING FIBER OPTIC COUPLED 
INPUT FIBER OPTIC FROM TARGE 




RETURN FIBER 
OPTIC TO 
VISAR 



figure 7. Drawing of imaging optical feedback 
sensor for VISAR. 



THEORY OF OPERATION 

In a typical experiment, a laser beam is focused to a 
small spot onto a target of interest. The reflected 
light is collected and routed to the interferometer 
cavity (figure 2). A dichroic mirror is inserted into 
the light collecting assembly to transmit the laser 
light and reflect the other wavelengths (This 
technique is valuable for viewing the target, allowing 
for precise alignment of the laser beam.). _ 



FOCUSINCMX)LLECTING LENS 
| TARGET y ^ ^TURNING MIRROR 




^/THRUHOLE 



t ^06N-16NLASKk ) 



§|||f^ICHROIC MIRROR 



'flllf • — FOCUSING LENS 



fVISAR 
CAVITY 



j 



FIBEROPTIC 



figure 2. Conventional method for collecting 
Doppler information from target 



- 310 - 



The return light, containing equally distributed S and 
P polarization components, is collimated and sent 
through the interferometer cavity (figure 3). A 
50/50 beamsplitter separates the light so that one 
beam travels through the glass "delay bar" and the 
other travels through air and an 1/8 wave retarder 
before recombining and producing an interference 
(fringe) pattern . 




tfSLATOR 



"INPUT LIGHT 
FROM TARGET 



MIRROR 
MV50 BEAMSPLITTER 



rLASS "DELAY BAR" 



figure 3. Fixed cavity VISAR schematic showing 
beam paths and piezo angular translator (PZAT). 
"Data" figures are photodetectors. 



In order to obtain quality fringe patterns, the image 
distances in both legs of the interferometer must be 
equal to within a few thousandths of a inch. If both 
legs of the interferometer are air, the measured 
distance would be equal. However, the refractive 
properties of glass make the image distance in the 
glass delay leg farther away than the image distance 
in the air reference leg. This relationship is defined 
by: 

x=h(l-l/n) 

where h is the delay leg length and n is the index of 
refraction. The distance the light has to travel in the 
delay leg is farther than the reference leg and the 
velocity of light is slower in glass than in air. Using 



the relationship from equation (1), the delay time x is 

given by: 

x=(2h/c)(n-l/n) 

where c is the speed of light in a vacuum. Using 

these relationships, the fringe count F(t) relates to 

target velocity u(t-t/2) as 7: 

XF(t) 1 
u(t- r/2) = — . 

in which A, is the wavelength of the laser light, Av/v 
is an index of refraction correction factor if a 
window is used, 6 is a correction factor with respect 
to wavelength for dispersion in the delay bar. 
Equation (3) may be manipulated to obtain the 
velocity-per-fringe (VPF)* constant for the 
interferometer. The VPF equation is: 

VPF i .-L 

2<l + Au/u) l + £ 

With these relationships, VISAR cavities with 
different sensitivities can be designed for optimal 
performance with regard to anticipated velocity 
versus Doppler resolution parameters. In any 
experiment, it is helpful to know that everything is 
operating correctly. Active feedback from the 
measuring instrument is a good method of assuring 
proper operation. The piezoelectric angular 
translator (PZAT) performs one such function by 
electrically moving a mirror in the cavity, effectively 
changing the cavity dimensions. When the cavity 
dimension changes by A/2, a 180° phase change 
occurs which effectively simulates the fringe record 
for a velocity change equivalent to one-half the VPF 
value. The return interference signal is monitored 
and the experimenter is now able to verify that the 
system is functioning correctly. 



* The VPF is a numerical constant unique to an 
interferometer cavity, typically given as mm/us or km/s. For 
instance, a cavity with a VPF of 1 would have an interference 
pattern of a 360° sine wave for a target accelerating to 1 
mm/us. 



- 311 - 



In figure 3 t one component of light passes through 
the 1/8 wave retarder twice, which makes it 90° out 
of phase with the other component. The polarizing- 
beam splitting cubes then separate the S from the P 
light and each beam is sent to a photo detector 
coupled to a digitizer. Recording two 90° out of 
phase signals produces an interference pattern that is 
a sinusoidal plot. Phase resolution of the signal is 
poor when the intensity of the sine wave is at a 
maxima or mimima. Making the two sinusoidal 
traces 90° out of phase insures that at any point in 
time one of the signals will be in a region of good 
resolution. Also, target acceleration or deceleration 
can be discriminated. During target acceleration, 
one signal pattern will lead the other by 90° and a 
deceleration will cause the opposite to occur. 

SOLID STATE VIS AR DEVELOPMENT 

The original intent for the development of the Solid 
State VISAR was for a portable "in house" tool that 
could be shared by several experimenters and used in 
the laboratory or in the field. At the same time the 
Defense Nuclear Agency (DNA) was interested in 
optical based instrumentation for use at (NTS). One 
particular experiment required an "up close" 
measurement of ground shock generated by the 
detonation. Since the nuclear event yields radiation 
and electro-magnetic fields, optical sensors are 
preferred because of their relative insensitivity to 
these phenomenon. Also, the sensors require no 
electricity which adds a greater margin of safety to 
personnel. The following are some of the 
requirements for the system to function: 

• Doppler measurement over kilometer-length 
fiber optics 

• Rugged system, operating on 120 VAC 

• Must run several days with no adjustments 

• Sensors must withstand mechanical and chemical 
abuse 

• Window and sensors must survive concrete 
encapsulation and measure ground shock a few 
meters from the detonation 

• Data acquisition bandwidth limited to 20 MHz 



EXPERIMENT DESCRIPTION 

Measurement of ground shock produced by the 
detonation is important because it contains valuable 
information used for analysis and modeling of the 
test. This ground shock measurement method uses a 
window, with sensors coupled by fiber optics to a 
VISAR cavity. The window is oriented towards the 
device and the entire area is filled with concrete. 
When the device detonates, a shock wave is 
transmitted through the concrete and into the 
window where the shock wave imparts a particle 
velocity in the window. The Doppler shift, which 
corresponds to the particle velocity in the window, 
is transmitted through the fiber optic to the VISAR 
cavity. The fringe data are converted to electronic 
signals and stored on digitizing oscilloscopes 
(digitizers). The mechanism for triggering the 
digitizers is a time of arrival (TOA) gauge. The 
TOA gauge is simply a fiber optic loop protruding in 
front of the window with a laser connected to one 
end and a photodetector attached to the other. 
When the shock wave breaks the fiber optic, laser 
light no longer enters the detector, it's output 
voltage drops, triggering the digitizers in VISAR. 

The characteristics of the window and the material 
transmitting the shock wave into the window must 
be known for accurate particle velocity 
measurements. The simplest way to use a window is 
to choose a material that has already been 
characterized. Unfortunately, the unusual laser 
wavelength and the large window thickness required 
for adequate recording time (40|as) did not allow 
previously characterized windows to be used. The 
material chosen for the window used in this 
experiment is Schott BK-7 glass, which has good 
broadband optical transmittance and is available in 
the 8" thick x 14" diameter required for the test. 
Several specimens of BK-7 as well as cored samples 
of the concrete were analyzed for their shock 
properties by impacting them into other known 
window materials, then into themselves using a gas 
gun as the target accelerator. The results from the 
analysis, commonly called a Shock Hugoniot, 



- 312 - 



determine whether the material is suitable for the 
shock pressure predicted for the event. The data 
also correlate the impedance mismatch at the 
grout/glass interface. The anticipated shock 
pressure for the NTS test, a few meters from the 
device, is « 70 IcBar in the grout. Figure 4 is a 
display of the type of plots for the shock pressure 
tests obtained using BK-7 as the impacted material. 
The BK-7 behaves like fused silica up to 90 kBar. 
Although BK-7 does not turn opaque at pressures 
above 90 kBar as fused silica does, the non-linear 
"shock-up" makes particle velocity correlations 
more difficult. 





































1™ 








-^"•tei 


K**r} 






J 








s~~lS 


K^ 






1 
























at 


£***• 






ft 
































































%M Tmm mm 


m LM LM ■.«• i.m MM 

TIME <US> 


figure 4. Data plot of BK-7 particle velocity versus 
time. 



The VISAR designed for the underground test 
contains a diode pumped Nd: YAG laser operating at 
1319 t|m wavelength, with a CW output of 160 
mW, and a 5 kHz, single frequency linewidth. This 
laser is a good choice for this system because it is 
stable, has low sensitivity to optical feedback, and 
the wavelength exhibits very low attenuation and 
high bandwidth in silica fiber optics. This high 
performance is due, in part, to the self cancellation 
of different wavelength-dependent dispersion effects 
that occur in fiber. The photodetectors in the 
system are comprised of indium-gallium-arsenide 
(InGaAs) photodiodes coupled to low noise/high 
gain operational amplifier circuitry. The peak 



sensitivity for these detectors happens to be at 1320 
rim wavelength, which not only affords greater 
sensitivity, but also increases the signal to noise 
ratio. The linear response range for the 
detector/amplifiers is DC to 125MHz with a linear 
response better than 3% and an output voltage of 
40mV/nW of light at 1320 rim wavelength (The flat 
linear response is critical for accurate data 
collection). 

One of the most critical parts of the system design 
was the fiber optic coupled sensors. Laser radiation 
is injected into a 50 micron graded index, multimode 
fiber optic connected to sensors which image the 
fiber optic onto the rear surface of the window. The 
return light reflected from the window's rear surface 
is collected and injected into a 100 |im step index, 
multimode fiber optic connected to the VISAR 
cavity. A redundant sensor is linked to the VISAR 
cavity in the event of damage to one of the sensors 
(figure 5). Correctly designing the optical train is 
paramount to attaining good signal strength that 
won't degrade under harsh environmental conditions 
(Obviously, after the sensors are encapsulated in 100 
foot thick concrete, re-aligning is impossible.). 
There was some concern about stress induced 
polarization after observing wildly fluctuating S and 
P ratios when the fiber optic was bent. These 
fluctuations in polarization will change the sine- 
cosine relationship critical to accurate data analysis. 
In most cases, this polarization problem occurs 
when highly polarized laser beams are injected in 
short lengths of fibers. The root cause is the mode 
structure is not fully mixed in the fiber optic and 
stressing the fiber optic redistributes these modes 
causing a change in the polarization (The laser 
output is single mode-single frequency and should 
not be confused with fiber optic propagation 
modes.). Since there is no room for error on the 
real test, three solutions are incorporated to remedy 
the problem: installing a mode scrambler that 
pinches the fiber optic into a serpentine shape, using 
a long fiber optic to mix the modes, installing a 
rotatable linear polarizer at the entry into the 



- 313 - 




TIME (microseconds) 
figure 6. Three of several TOA gauges 



interferometer cavity. The new modifications solved 

the problem. 

The sensors perform three functions; collimating and 



1 

0.5 



| -0.5 

-1 

-1.5 

•2 
0.I 

figure 7. 
90°out< 




, r ^~- 


h 


/ 


j 


/ 


■ ) 


i^ 


l 


955.06 •* 
►54. Wm 

m 

i.i.i 


95 0.95 0.95 0.96 0.96 
Tlme(ms) 

Raw, unreduced data. Notice the 
f -phase signals. 



focusing the laser radiation onto the rear surface of 
the window, collecting the return light, and injecting 
the light into the return fiber optic. The sensor 
configuration is similar to figure 1 except that the 



camera is omitted, since there is no need to see the 
rear window surface. 



TOA GAUG ES 
I = 



FIBER OPTIC LOOP 



SHOCK 



WAyE ■ 




CKUP SENSOR ^N, 

ER OPTIC SENSOR 



8"x 14" BK7 GLASS 
MIRRORED COATING 



FOCUSING 
OPTICS 

SOLID STATE LASE 



50 um fibe r 
;>ptic cable 



iobi 



100 um fiber 
optic cable 





DETE 7TOR 



- VIS AR CAVITY 



figure 5. Diagrammatic layout of the window, 
sensors, time of arrival (TOA) gauges, VISAR 
cavity, and laser. 



Although the primary function of the TOA gauge is 
to trigger the digitizers, utilizing a TOA array 
provided additional shock data. Several gauges 
were placed around the window with the tips of the 
TOAs staggered to intercept the shock wave front 
at different points in time (figure 6). As the shock 
front breaks the TOAs, the digitizers record the 
time-of-break that is then correlated to shocks 
arrival time and velocity. 



EXPERIMENTAL RESULTS 

The VISAR and TOA gauges performed with 
strong, clean signals recorded on all instrumentation 
channels. The Doppler information, in unreduced 
form, is shown in figure 7. The signal strength is 
lVolt peak to peak which is well above the noise 
floor that has, in the past, caused difficulty in 
accurate data reduction. The 90° phase relationship 
between the two traces indicates the stress induced 
polarization problem has been cured. 

Figure 8 shows the reduced data. There is an 
impedance mismatch between the BK-7 and the 
grout but the shock Hugoniot for these materials is 



- 314 - 



known and was used in calculating the final particle 
velocity. The peak recorded particle velocity was on 
the order of 0.6 mm/us, which corresponds to a 
pressure at the interface of 85 kBar. 



figure 8. Reduced data showing particle velocity 
versus time 



The results indicate that the yield of the device was 
greater than expected (60 kBar). The choice of 
BK-7 was fortunate because if fused silica was 
available and used, the data would have been lost. 
(Fused silica is known to become opaque at 
pressures above * 82 kBar.) The BK-7 is apparently 
able to withstand slightly greater shock pressures 
before going opaque and at 1/3 the cost, it is 
significantly less expensive. 



ACKNOWLEDGMENTS 

This work was performed at Sandia National 
Laboratories, Albuquerque, NM for the United 
States Department of Energy under Contract DE- 
AC04-76DP0089. This project was truly a team 
effort and the authors gratefully thank the following 
people for their participation: 
Leonard Duda- financial and technical design 
support, John Matthews and Richard Wickstrom- 
software support, Larry Weirick, Mike Navarro, 



Heidi Anderson- BK-7 characterization & gas gun 
support, Bill Brigham, Theresa Broyles, Dan 
Sanchez- Electronic design & fabrication, Dan Dow- 
mechanical fabrication, Terry Steinfort, Robert 
Vasquez- mechanical design & installation, and 
William TarbelU shock analysis. A special thanks to 
Lloyd Bonzon for supporting the concept in its 
infancy. 

REFERENCES 

1. L.M. Barker and R. E. Hollenbach, Laser 
Interferometer for Measuring High Velocities of 
Any Reflecting Surface. Journal of Applied 
Physics 43:11 (Nov 1972) 

2. W.F. Hemsing, Velocity Sensing Interferometer 
(VISAR) Modification. "Review of Scientific 
Instruments 50:1 (Jan 1979). 

3. IE. Kennedy, Org 7130, private communication. 

4. KJ. Fleming, O.B. Crump Jr., Portable. Solid 
State VISAR . SAND92-J36L April, 1992 

5. O.B. Crump Jr., PL. Stanton, W.C. Sweatt, The 
Fixed Cavity VISAR. SAND92-0162 (March 
1 992). Sandia National Laboratories, 
Albuquerque, NM. 

6. K.J. Fleming, Fiber Optic Coupled Probe for 
Versatile Interferometry . Department of Energy 
Patent Disclosure SD-5034, S-74-181 

7. L.M. Barker and K.W. Schuler, Velocity 
Interferometry System . Journal of Applied 
Physics. 45,3692(1974). 



- 315 - 






DEVELOPMENT AND QUALIFICATION TESTING OF A 
LASER-IGNITED, ALL-SECONDARY (DDT) DETONATOR 

Mr. Thomas J. Blachowski Mr. Darrin Z. Krivitsky 

CAD R&D/PIP Branch CAD Weapons/Aircraft Systems Branch 

Indian Head Division Indian Head Division 

Naval Surface Warfare Center Naval Surface Warfare Center 

Indian Head, MD 20640 Indian Head, MD 20640 

Mr. Stephen Tipton 

B-1B Systems Engineering Branch 

Oklahoma City 

Air Logistics Center 

Tinker AFB, OK 73145 

Abstract; 

The Indian Head Division, Naval Surface Warfare Center (IHDIV, NSWC) is 
conducting a qualification program for a laser-ignited, all-secondary (DDT) 
explosive detonator. This detonator was developed jointly by IHDIV, NSWC and 
the Department of Energy's EG&G Mound Applied Technologies facility in 
Miamisburg, Ohio to accept a laser initiation signal and produce a fully 
developed shock wave output. The detonator performance requirements were 
established by the on-going IHDIV, NSWC Laser Initiated Transfer Energy 
Subsystem (LITES) advanced development program. Qualification of the 
detonator as a component utilizing existing military specifications is the 
selected approach for this program. The detonator is a deflagration-to- 
detonator transfer (DDT) device using a secondary explosive, HMX, to generate 
the required shock wave output. The prototype development and initial system 
integration tests for the LITES and for the detonator were reported at the 
1992 International Pyrotechnics Society Symposium and at the 1992 Survival and 
Flight Equipment National Symposium. Recent results are presented for the 
all-fire sensitivity and qualification tests conducted at two different laser 
initiation pulses. 



- 317 - 



Introduction: 



The INDIV, NSWC 
Devices (CADs), and Ai 
The CAD Engineering Di 
development and engine 
sources , energy transm 
CADs, and their asscci 
three services. The C 
policy and conducts se 
power signal transmiss 
stores /weapons deliver 
In addition, administr 
Improvement Programs ( 
for CADs is performed 
for cartridges, CADs, 
authority. 



s the lead service for cartridges, Cartridge Actuated 
rcrew Escape Propulsion Systems (AEPS) for all services. 
vision at IHDIV, NSWC manages and implements the 
ering functions ir. the fields of ballistic power 
ission systems, and control devices such as cartridges, 
ated signal or energy transmission subsystems for the 
AD Engineering Division also recommends and establishes 
rvice qualification for cartridges, CADs, and related 
ion systems for aircrew escape and survival and 
y and deployment for the Department of Defense (DoD) . 
at ion of research, design, development, and Product 
PIP) for ignition devices and -electro-explosive devices 
by the Division. Development of design specifications 
and related ballistic systems is maintained under this 



Background t 

The IHDIV, NSWC has pursued development of LITES for these applications. 
LITES utilizes chemical f lashlamps to generate light which is amplified by a 
laser rod, focused into fiber optic transfer lines, and delivered to optically 
initiated pyrotechnic output devices. Proof-of-Principle tests and the 
exploratory development program (13th International Pyrotechnics Society 
Proceedings - 1988) were completed. An advanced development program was 
conducted to miniaturize the laser housing while maintaining the proven 
neodymium doped phosphate glass rod and commercially available flashlamp 
concept. The miniaturized laser (Figure 1) is a mechanically actuated device 
which generates sufficient energy, when delivered through a bundle of fiber 
optic lines, to initiate an optical output device. The LITES advanced 
development program has designed optical output devices which produce 
ballistic pressure or a shock wave output (detonator) as required for a 
specific aircrew escape system application (18th International Pyrotechnics 
Society Symposium - 1992). FLashlam-s 

f-m — -**'*^' 

| ussa ROO 

PUCSHATS GLASS 




1.15 INCnSS 
INITIATION m£C*AVSm 

F:S=K CrTC LINs 

Figure 1: LITES Mechanically Actuated Laser Assembly 



- 318 - 



As part cf th 
Mound AppL ied Techr. 
(Table l") for the I 
to-de to nation trans 
wavelength input ar. 
commercially aval la 
requirements , alone 
a Shielded Mild Det 
independent cf the 
specifically for us 
detonator to be use 
line configuration 
technology required 
design goal through 
Pyrotechnics Societ 






v Svmc 



:t~:s :::;: 


art , t r. 


rough 


a. set! 


es cf discussions with EG&G 


^^95 ^ 1 r.Z Z 


V N S ** 


C est 


s c 1 l s h e 


d the following requirements 


r— igr. i «s - 


deccr.a 


tor. 


Soecif 


icaily, this deflagration- 


(DO? j cev 


ice ( F 


igure 


2 ) wou 


id accept a 1.06 pm laser 


culd cr.lv 


cental 


r. set 


tndary 


explosives that are 


from save 


ral sc 


urces 


in the 


United States. These 


th the win 


— '-j w a . . 


c gen 


e rat ion 


of an output equivalent to 


ting Core: 


(SHDGj 


tip, 


allows 


the optical detonator to be 


er attic s 


ignal 




tission 


system. Although developed 


ith LUES, 


e f f c r 


t s we 


re take 


n to permit this optical 


%- a variet 


v cf e 


r.ercv 


culse 


duration and fiber optic 


er ignitic 


n s v s t 


ems . 


x n a ^ >-i 


ition, transition of the 


manuf accu 


re the 


detc 


r.ator t 


o industrv was held as a 


this deve 


Icemen 


t pro 


g r am (1 


3th International 


vT.tosiu" - 


1992 ) 









Detonator Requirements 


• 


All secondary explosive device 


• 


Flat window configuration 


• 


Shielded Mild Detonating Cord (SMDC) tip output 


• 


Hermetic all v sealed 


• 


Structurally sound, all Reactants contained 


• 


No proprietary components 


• 


Transition government developed technology to 
industry for competitive procurement 



Table 1: Laser-Ignited Detonator Design Requirements 




Figure 2: Laser-Ignited, All Secondary (DDT) Detonator 



- 319 - 



Test Program Preparation: 

Definitions: 

Prior to establishing the qualification test matrix and performance 
requirements for the laser-ignited detonator itself, it was necessary to 
define the components and the parameters involved in the overall aircrew 
escape system. For this paper, an Ordnance Initiation System is defined as a 
"^system consisting of three distinct components: (1) a signal generator and 
controller which is capable of establishing an initiation pulse of sufficient 
intensity at the required time interval, (2 ) a signal transmission system 
capable of transferring the pulse to all output devices within the envelope of 
the application, and (3) output devices (initiators or detonators) which 
either perform the required work function directly or initiate a second in-" 
line device which performs the function. As previously described, these 
output devices include items such as initiators, squibs, gas generators, 
electro-explosive devices, thermal batteries, cutters, and detonators. An 
example of an aircrew escape system ordnance initiation system is shown in 
Figure 3 . 




> 



Figure 3: Generic Aircrew Escape System 

During the evaluation of an ordnance initiation system, the parameters 
of the output devices must be clearly identified. Several existing government 
specifications clearly define the "all fire" energy of an output device and 
the "no fire" energy of an output device. The "all fire" energy level of an 
output device is defined as the minimum amount of energy or power required to 
initiate that output device in its final configuration with a reliability of 
0.99 at a 0.95 confidence level. The "all fire" energy level will be 
determined by any suitable test method consisting of a sample size of not less 
than 20 output devices. This quantity is not defined in the existing 
specifications; however, a statistically significant sample size must be 
tested to verify the stated energy levels. 

Threshold energy levels, a 50% initiation energy level, or a reliability 
of 0.9999 at a 0.98 confidence level are technically very useful values and 
may be required for a specific application. Any of these values; however, 
cannot be identified as the "all fire" energy level of an output device. 

Conversely, the "no fire" energy level of an output device is the 
maximum amount of energy or power which does not initiate the output device in 
its final configuration within five minutes of application. At this 
initiation level, less than 1.0 per cent of all output devices at a level of 
confidence of 0.95 can actuate. Again, the "no fire" energy level will be 
determined by any suitable test method consisting of a sample size of not less 
than 20 output devices. 



- 320 - 



Current Specifications: 



d general specifications in place to specifically address 
ultimately implementation of laser initiation system 



There are no 
qualification and ultimately implc 

technology in the U. S. Department of Defense, the Department of Energy, and 
the National Aeronautics and Space Agency (NASA). Much discussion has taken 
place over the past years as to which existing specification or series of 
specifications are most applicable to this new technology. A partial list has 
"been compiled of the specifications most often mentioned in these discussions 
(Table 2) . 



•--^-Specification \ | -y^-:. Title 


MIL-STD-1316 


• Safety Criteria for Fuze Design 


MIL-STD-1512 


• Design Requirements and Test Methods for 
Electrically Initiated Electroexplosive 
Subsystems 


MIL-STD-1576 


• Safety Requirements amd Test Methods for 
Space Systems Electroexplosive Subsystems 


MIL-E-83578 


• General Specification for Explosive Ordnance 
for Space Vehicles 


MIL-C-83125 


• General Design Specification for Cartridges 
and Cartridge Actuated/Propellent Actuated 
Devices 


MIL-C-83124 


• General Design Specification for Cartridge 
Actuated Devices/Propellent Actuated Devices 


MIL-D-81980 


• General Design Specification for Design and 
Evaluation of Signal Transmission Subsytems 


HIL-I-23659 


• General Design Specification for Electric 
Initiators 


MIL-D-21625 


• Design and Evaluation of Cartridges for 
Cartridge Actuated Devices 


MIL-D-23615 


• Design and Evaluation of Cartridge Actuated 
Devices 



Table 2: Current Specifications for Laser Technology Implementation 



Specification Selection Process: 

To successfully conduct and complete development and qualification 
testing on the laser-ignited detonator and to ultimately implement this device 
and other laser initiation system technology into next generation aircrew 
escape systems, three different approaches have been identified (Table 3, next 
page). 



- 321 - 



APPROACH ! ADVANTAGES 


DISADVANTAGES 


NEW 

SPECIFICATION 


• GENERAL FORMAT 

• TECHNOLOGY SPECIFIC 


• LENGTHY APPROVAL PROCESS 

• NEW REQUIREMENTS 

• TECHNOLOGY NOT MATURE 


SYSTEM 
SPECIFIC 
DOCUMENT 


• SPECIFIC REQUIREMENTS MET 

• WRITTEN TO SCHEDULE 

• ?JZ/?SQ DOCUMENTS PREPARED 


• NOT GENERAL 

• VERY DETAILED 


EXISTING 
SPECIFICATION 


• GENERAL FORMAT 

• NO NEW REQUIREMENTS 
•- PRECEDENCE 


• DATED REQUIREMENTS 

• MUST 3E AMENDED 



Table 3: Specification Approach Advantages/Disadvantages 

The first approach is to generate a new specification for laser 
initiation sysiem components . The general format of this type of 
specification will allow implementation of the technology on a wide variety of 
platforms. The technology definitions included in this specification will 
allow Program Managers and design engineers the ability to better compare 
alternatives during the preparation of Trade Studies, program plans, etc. The 
required testing and reporting will better establish this technology baseline 
that will serve all users. 

The disadvantages of this approach are severe. Laser initiation 
technology is rapidly advancing to new levels. What was thought to be a 
technical barrier a few years ago has changed. As these continued 
advancements occur, there has not been sufficient time for the "off-the-shelf 
components to mature. The baseline has been moving. Writing a specification 
without fully identifying this baseline is very difficult. Once an agency has 
officially undertaken the specification writing task, there is a well defined 
and lengthy approval process through any Department of the Government. With 
the constantly changing baseline of the technology and the lengthy approval 
process, this approach is unattractive. 



The second approach is to allow Program Managers to 
specifications for a single system. The advantages of th 
numerous. The Program Managers have a "hands on" feel of 
their platform. Specialized needs, power requirements, s 
performance among other parameters would be contained in 
Potential suppliers then have a defined goal to design an 
technology alternative as a solution and all the pre-sele 
reporting needs are highlighted. Competing solutions, tr 
overall program technical risk are clearly outlined to th 
This overall approach will be conducted for every system; 
of assets allocated (management time, engineering assets) 



prepare 
is approach are 

the requirements for 
afety margins, output 
this single document. 

individual 
ction testing and 
ade studies, and the 
e Program Managers. 

however, the amount 

is great. 



This leads directly into the disadvantages of this 
detailed system specific document requires an investment 
time and resources. The amount of technical detail in t 
depend on the individual program office or on the contra 
their preparation. Issues such as competitive procureme 
technology or sole sourcing the procurement to a particu 
life of the platform must be addressed at this step. If 
document becomes too detailed, sole source procurement o 
very likely. Upon completion of these documents, they rr. 
to other platforms. Some requirements may transition to 
well; however, most will not. 



approach. The very 

of Program Management 
hese documents will 
ctor assigned with 
nt of the selected 
lar vendor for the 

the system specific 
r its variations, are 
ay not be applicable 

a general system very 



- 322 - 



The third approach is to utilise existing specifications to implement 
laser initiation technology into current and planned platforms. There are 
several advantages to this approach. Existing specifications are written to a 
general format thus allowing all alternative technologies to design 
engineering solutions. The previous test requirements for each platform are 
well defined and established. Potential vendors for a specific platform have 
a defined baseline of past knowledge to build upon. Precedence has been 
"established by the Program Managers utilizing these specifications. There are 
no new technical limitations for implementing laser initiation technology onto 
any platforms. 

The disadvantages of this approach include the time dated nature of all 
the listed specifications (Table 1), and they must be amended somewhat to 
address new technical concerns. The listed specifications were written to 
address the general design issues of that time and, even including amendments 
and re-issuances, do net address some of the attributes of laser initiation 
technology. Minor modifications to these specifications to address these 
special attributes allow these documents to govern implementation of a new 
technology onto current and future DoD platforms. 

Existing Specifications Selected: 

Following this process, the third approach of selecting existing 
specifications was chosen to allow for implementation of laser initiation 
system technology into DoD applications. The specifications selected by 
IHDXV, NSWC are as follows: 

(1) the signal generator and controller (Laser Assembly) will be 
governed by MIL-C-83124, 

(2) the signal transmission system (STS) will be governed by 
MIL-D-81980, and 

(3) the optical output devices (Initiators and detonators) will be 
governed by MIL-C-83125. 

These documents were selected because MIL-C-83124 and MIL-C-83125 apply 
to energy sources and ballistic devices (cartridges and CADs) for all three 
services. This tri-service approval is very attractive in making the 
specification requirements as general as possible while still meeting a DoD 
baseline. For the STS, MIL-D-81980 was selected. This document is a 
Department of the Navy specification but has precedence in testing energy 
transfer systems for the other services. 

The laser-ignited detonator is an optical output device and the 
Qualification Test Procedure (QTP) to allow for qualification and 
implementation of this device was written against the requirements specified 
in MIL-C-83125. 

Qualification Test Procedure: 

A QTP was prepared by IHDIV, NSWC to detail all environmental test 
conditions and output performance requirements for the laser-ignited 
detonator. As previously described, this procedure governs only the laser- 
ignited detonator; the laser signal generator (Laser Assembly) and the fiber 
optic signal transmission system will not be subjected to these environmental 
tests. All of the environmental tests and required number of detonators are 
shown in Table 4 (next page). The test matrix requires 161 detonators to 
successfully demonstrate this technical concept. 



- 323 - 



Environment*] Tcsi / 

Quantity 


4 J « 


6 


9 9 

i 


9 


9 


12 


12 


6 


9 


30 


10 


30 


Visual Inspection 


4 


6 


6 


9 


9 


9 


9 


12 


12 


6 


9 


30 


10 


30 


X-Riy <fc n-Rjy Inspection 


4 


6 


6 


9 


9 


9 


9 


12 


12 


6 


9 


30 


10 


30 


Dry Gas Leakage 


4 


6 


6 


9 


9 


9 


9 


12 


12 


6 


9 


30 


10 


30 


Noa-Electric Initiation 
(-90° F) 


4 




























40* Drop Test 




6 


























6* Drop Test (70° F) * 






6 
























• Shock 








9 


9 


9 


9 
















• TSH&A 










9 


9 


9 
















• Low Temp. Conditioning 












9 


9 
















• VTb ration 














9 
















Cook-Off 
















12 














Hifh Temp. Exposure (70* F) 


















12 












Salt Fog (70* F) 




















6 










Hi$h Temp. Storage (200 # F) 






















9 








Functional Low Temp. 
(-65* F) 
























30 






Functional Ambient Temp. 
(70* F) 


























10 




Functional Hi^h Temp. 
(225* F) 




























30 



Table 4: Qualification Test Matrix for Laser-Ignited Detonator 



Table 4 Notes: 



(1) TSH&A is defined as Temperature, Shock, Humidity, and 
Altitude Cycling. 

(2) The temperature in parentheses is the functional firing 
temperature of the detonators, 

(3) The "•" denotes SEQUENTIAL TESTING which means the nine 
vibration detonators were subjected to all four 
environments and functionally tested; three detonators 
at -65° F, three detonators at 70° F, and three 
detonators at 200°F. The Low Temp. Conditioned 
detonatorB were also subjected to TSH&A cycling and 
Shock prior to functional tests as described above. 

(4) For the 40' Drop and Cook-Off Tests, no functional 
test firings are required. The detonators must survive 
these environments and be safe to handle and discard. 



- 324 - 



Successful completion of this QT? demonstrates the laser-ignited 
detonator concept. Additional testing will be required prior to release of 
this device to service use. The application specific locking connectors (both 
from the fiber optic transmission system to the detonator and from the 
detonator to its work performing device) must be demonstrated through a 
similar environmental test series prior to aircrew escape system 
implementation. 

Short Pulse Qualif ication Test Results: 

Defining the specific laser energy pulse, its duration, and the 
configuration of the energy delivery system was performed at this time. 
Driven by a specific aircrew escape system application, the viability of a 
microsecond(s) long pulse duration was considered. Based on the 
recommendations from EGSG Mound and by this Activity's research, the laser- 
ignited detonator was capable of being successfully initiated with a pulse of 
this duration. A 150-microsecond pulse duration delivered through a hard clad 
silica (numerical aperture of 0.37) fiber optic line was established as the 
initiation condition. A 20 unit Neyer threshold test* was conducted to 
determine the all-fire energy of the laser-ignited detonator. Based on these 
test results, the 0.99 reliable at a 0.95 confidence interval energy value was 
determined to be 131.3 milli joules in this configuration. The diagnostic 
equipment was verified for this conf icuration (Figure 4). To further insure 
initiation of the detonator in this configuration, the following values were 
used to begin the qualification testing: 

• 150 millijoules of laser energy 

• 150 microsecond pulse duration 

• 200 micron fiber (NA « 0.37) 

Nd:YAG Laser Pulse 



0.10 



0.08 



« 0.06 



o 

M 


0.04 


> 

J5 


0.C2 


GC 


0.00 




•0.C2 




•50 50 100 

Time - microseconds 



150 



200 



Figure 4: 150 Milli joule Laser Energy Pulse from Quanta ray DCR-2A Laser 

Ten detonators were selected from the Functional Ambient Temperature 
group of the QT? to begin the testing (Table 5, next page). The first six 
detonators successfully initiated and met all the performance requirements. 
The seventh detonator did not initiate. After a series of discussions, IHDIV, 
NSWC and EG&G Mound representatives agreed to continue the testing. The 
eighth and ninth detonators functioned as designed. The tenth detonator did 
not initiate. At this point, the qualification testing was halted. 



1 - "More Efficient Sensitivity Testing" 
Applied Technologies - October 20, 1989. 



Barry T. Neyer, MLM-3609 EG&G Mound 



- 325 



Tcrt 



10 



Function 
Temp* ram re 



70* F 



70* F 



70° F 



70° F 



70* F 



70° F 



70° F 



70° F 



70° F 



70' F 



Pre-Shc* Tent* 
(ml) 



151.9 ± 1.3 



151.2 ± 0.4 



149.5 ± 0.3 



150.3 ± 0.6 



149.7 ± 2.2 



149.7 ± 1.7 



149.6 ± 1.4 



152.3 ± 0.4 



148.7 ± 0.6 



150.9 ± 2.6 



Function? 



YES 



YES 



YES 



YES 



YES 



YES 



NO 



YES 



YES 



NO 



Pulae DuT*Lk>n 



150 



150 



Function Time 
(pscc} 



55.5 



77.5 



150 



58.0 



153 



50.5 



150 



63.5 



153 



71.5 



150 



150 



60.0 



149 



73.0 



156 



Indent 
Cm.) 



0.058 



0.052 



0.051 



0.055 



0.055 



0.053 



0.051 



0.051 



Table 5: Short Pulse Laser-Ignited Detonator Test Results 

Short Pulse Failure Investigation: 

A complete failure investigation was undertaken to determine the cause 
of the non-initiation of these detonators and to establish engineering 
solutions to eliminate these non-initiation causes from the design concept. 
This was a wide ranging investigation which included a re-assessment of the 
window design and status during the testing, the energetic material post-test 
condition, and the diagnostic test equipment itself. 

The fiber optic line delivering the required energy pulse was examined. 
A standard SMA-905 connector was used to link the line to the detonator. This 
connector centers the fiber optic line in a stainless steel sleeve with epoxy 
filling the surrounding gap. One potential failure cause was that if this 
epoxy slightly covered the face of the fiber or at the tip of the fiber, the 
epoxy was vaporized as the energy pulse exits the fiber. This phenomena would 
block the laser energy from the window and result in a greatly decreased 
amount reaching the energetic material, inducing non-initiation of the 
detonator. 

A second cause involving damage to the window was also investigated. 
During detonator manufacture, optical transmission of the window was tested 
prior to loading of the energetic material. This detonator design requires a 
contact between the fiber optic line and the window. Due to this design and 
the optical transmission testing, surface damage occurred to the window. 
These damage patterns were identified for all detonators and these patterns 
were grouped into eight categories: . flawless windows, small pits in the 
center of the window, small pits in the center and damage outside the center 
of the window, very light scratches and small dots on the window, deep 
scratches, film coating on the window, a deep inclusion in the window, and 
small center dots in the window with extra debris. Either or both of these 
investigated causes could result in non-initiation of the detonator. 



- 326 - 



Based on this completed- failure investigation and the assets available 
to continue this development and qualification program, the overall system 
initiation conf igura tier, was re-addressed. A question was posed to the 
aircrew escape system and aircraft system designers, "Could a system be 
designed, within existing aircraft parameters, to generate and support a laser 
pulse duration of 12 milliseconds?" The response from these designers was 
that a 12 millisecond long pulse could be implemented to resolve the 
demonstrated failure pattern. 

Long Pulse Qualification Test Results : 

Re-establishing the detonator conceptual design initiation pulse 
duration was the primary solution to the non-initiation experienced during the 
short pulse qualification testing. This longer duration lowers the power 
density of the pulse and greatly lessens the potential for laser induced 
damage in the window and/or the fiber optic line. In addition, lowering this 
power density and increasing the pulse duration renders slight imperfections 
in the window less critical to successful initiation of the detonator. This 
pulse duration allows the laboratory designed and developed detonator to be 
demonstrated as rugged enough for field applications* No enhancements, such 
as re-polishing the windows to reduce surface damage, were performed on any of 
the environmentally stressed detonators. 

As for the transmission system itself and the diagnostic test equipment, 
to further reduce the possibility of inducing a non-initiation, a glass/glass 
fiber optic line was selected (with a numerical aperture of 0.22). The SMA- 
905 end connector was assembled into this line utilizing a minimum of epoxy 
that was held away from the fiber tip itself* A 20 unit Neyer threshold test 
was conducted in this configuration to determine the 0.99 reliable at a 0.95 
confidence interval all fire energy of the detonator. Based on this test 
series, the following values were used for the long pulse qualification 
testing: 

• 132.8 millijoules of laser energy 

• 12 millisecond pulse duration 

• 200 micron fiber (NA « 0.22) 

The 132.8 millijoule, 12 millisecond laser energy pulse was confirmed 
through the diagnostic test equipment as before. The Function Time (defined 
as the time from laser pulse initiation to the output shock wave impacting a 
detector at the back of the test fixture) of the detonators was recorded and a 
minimum of 0.040 inch indent was established as the detonator output 
requirement. A total of 131 detonators were functionally tested using the 
initiation configuration and the results are grouped by environmental test 
condition (Table 6, next page). An additional 18 detonators successfully 
completed the QT? requirements without undergoing functional testing (the 40' 
Foot Drop Test and the Cook-off detonators). Also, 12 detonators that had 
undergone environmental conditioning were functionally tested for information 
only (Table 7, second page). Using the long pulse configuration, the laser- 
ignited detonator did not achieve all of its design goals for this QTP. 



- 327 - 



■■"■■ Environmental 


Function 
Temp. (*F) 


I Required 


/ Successful 


Function 
Tune (msec.) 


Indent 


Other 
Results 


NON-ELECTRIC 
INmATION 


-90* F 


4 


4 


3.59 ± 0.89 


0.053 ± 0.003 


NONE 


6 FOOT DROP 


70" F 


6 


6 


4.54 ± 2.23 


0.053 ± 0.002 


NONE 


SHOCK 


-65* F 

70' F 

200* F 


3 
3 
3 


3 
3 
3 


5.55 ± 0.68 
3.21 ±0.77 
6.33 ± 0.42 


0.054 ± 0.002 
0.054 ± 0.001 
0.050 ± 0.002 


NONE 
NONE 
NONE 


SHOCK, TSH&A 


-65° F 

70* F 

200* F 


3 
3 

3 


2 
2 

2 


7.57 ± 1.42 
6.36 ± 0.23 
6.99 ± 1.18 


0.048 ± 0.001 
0.O48 ± 0.004 
0.O43 ± 0.002 


(0.027) 

(0.025) 

(0.026) {0.041} 


SHOCK, TSH&A, 
LOW TEMP. 


-65* F 

70*F 

200* F 


3 
3 
3 


3 

1 
2 


6.16 ± 0.21 
6.85 ± 0.84 
6.41 ± 0.48 


0.050 ± 0.0W 

0.046 
0.045 ± 0.001 


NONE 

(0.016) (0.031) 

(0.025) 


SHOCK, TSH&A, 
LOW TEMP., 
VIBRATION 


-65* F 

70* F 

200* F 


3 
3 
3 


1 
2 
3 


6.18 ± 1.00 
3.30 ± 1.02 
7.59 ± 1.23 


0.042 
0.055 ± 0.000 
0.O4S ± 0.004 


(0.031) (0.033) 

1 

NONE 


SALT FOG 


70' F 


6 


6 


3.65 ± 1.06 


0.051 ± 0.004 


NONE 


HIGH TEMP. 

STORAGE 


200* F 


9 


8 


6.60 ± 1.78 


0.050 ± 0.003 


(0.031) 


LOW 

TEMPERATURE 


-65 # F 


30 


30 


3.62 ± 1.12 


0.051 ±0.003 


NONE 


AMBIENT 
TEMPERATURE 


70* F 


10 


10 


2.83 ± 0.43 


0.053 ± 0.002 


NONE 


HIGH 

TEMPERATURE 


225* F 


30 


25 


7.71 ± 1 JO 


0.051 ± 0.003 


(0.025) (0.025) 
(0.033) (0.025) 
(0.024) {0.042} 



Table 6: Long Pulse Laser Ignited Detonator QTP Results 



Table 6 Notes: 



(1) The Other Results indicated in " ( ) M are 
unacceptable indents. They are below the QTP mandated 
0.040 inch indent. 

(2) The Other Results indicated in "{ } M are marginal 
indents. These very closely achieve the 0.040 inch 
indent . 

(3) The n l n indicates an initiation failure. 



- 328 - 



[ Environmental 


Temp. CD 


# Rtqurrcd 


# Successful 


Function 
Time (msec.) 


I~dcn! 

Go.) 


Other 
Resuhs 


COOK-OFF 

SURVIVORS 


375 °F / 70* F 
400* F / 73* F 


3 
1 


2 



11.85 ± 3.21 
5.36 


0.049 ± 0.001 


(0.016) 
(0.011) 


HIGH TEMP. 
EXPOSURE 


325° F/70° F 
300° F / 70° F 
n 75* F ■ "0* F 


3 

2 

3 







12.32 ± 2.66 
8.76 ± 1.36 
13.18 ± 0.54 




ALL £ (0.015) 
(0.025) (0.022) 
(0.021) (0.023) 



Table 7: Long Pulse Laser Ignited Detonator Non-QTP Results 



Table 7 Notes: 



(1) The Other Results indicated in " ( ) " are 
unacceptable indents- They are below the QT? mandated 
. 040 inch indent . 

(2) The "2 M indicates an initiation failure. 



Long Pulse Failure Investigation: 

At this writing, the long pulse failure investigations are underway. 
There are two separate investigations being conducted by IHDIV, NSWC and EG&G 
Mound personnel: the first is to determine the cause of the two non- 
initiations, and the second is to determine the apparent temperature and/or^ 
temperature cycling effect on the detonator and its not consistently achieving 
the minimum 0.040 inch indent into an aluminum dent block. 

Determining the cause of the non-initiations is the first priority. Of 
the 143 functional detonator tests completed, there were 2 non-initiations. 
Both of these detonators had been subjected to elevated temperature 
environments (one during the TSH&A cycling had seen 160° F and the other 
during High Temperature Exposure had seen 275° F for a period of 12 hours). 
The detonators that had passed the indent requirement exhibited longer 
function times after being subjected to elevated temperatures. Some of the 
detonators in the non-QTP test series had even functioned after the 12 
millisecond laser pulse was completed. For the detonators subjected to cold 
environments, the function times are somewhat faster and all of these indents 
are acceptable. The diagnostic test set-up and operator handling procedures 
are also under review as part of this failure investigation. All of this 
information is being evaluated to determine the cause of these two non- 
initiations. 

The second investigation to determine the lack of sufficient indent into 
the dent block is also of great importance. Obtaining the 0.040 inch indent 
demonstrates this HKX, laser-ignited detonator is a one-fcr-one replacement 
candidate for the widely used SMDC lines and output tips which use HNS 
(Hexanitostilbene) as their energetic material. Demonstrating an identical 
indent for this laser-ignited detonator will greatly reduce the number of 
future tests required to assure this one-for-one replacement in all fielded 
applications. To begin this investigation, a record of the post-test 
detonator column condition is being compiled. For tests that achieved the 
0.040 inch indent, the column had fragmented or blossomed outward. In some 
cases, this expansion had not fragmented the metal column, just widened it 
slightly. And, in some other cases, the output column of the detonator 
remained the same size. Several theories are being explored to explain these 
test results. Once these theories are proven through additional testing, 
engineering solutions can be implemented to the detonator design concept to 
eliminate the potential of low indents. 



- 329 - 



Conclusions : 

This paper has presented a new laser-ignited detonator concept developed 
jointly by IHDIV, NSWC and EG&G Mound. Development and qualification methods 
for this new technology and new device have been presented using existing 
military specifications to establish the acceptance requirements. Diagnostic 
test equipment development, set-up, and specialized operating procedures were 
designed to demonstrate the performance of the detonator. Two Neyer 
"sensitivity test series were conducted to establish the "all fire- energy 
level. Two different initiation systems (different pulse durations, all fire 
energy levels, and connector interfaces) were investigated during this 
program. The laser-ignited detonator design was demonstrated as feasible 
within the system constraints. The concept^ is not completed. The reasons for 
non-initiation and the low indent results must be identified and resolved 
before this device is subjected to further system tests. Through the on-going 
failure investigations, solutions to these shortfalls are seemingly 
attainable. Upon implementation of these solutions, this detonator will be 
subjected to a final test series. Successful completion cf this delta 
qualification test series will allow the detonator to be released for field 
applications including aircrew escape systems. 

Biography; 

Mr. Thomas J. Blachowski has held his present position as an Aerospace 
Engineer in the CAD Research and Development/Product Improvement Program* 
Branch at IHDIV, NSWC for the past 8 years. Ke has been directly involved in 
LITES, laser initiation, and laser detonator development efforts since 1988. 
In addition, he has managed the Cartridge Actuated Device (CAD) Exploratory 
Development and Advanced Development programs since 1989. Mr. Blachowski 
received his Bachelor of Science degree from The Ohio State University in 
1985. 



- 330 - 




EXCELLENCE 
BY DESIGN 



ENGINEERING DIRECTORATE 




NA 



Lewis Research Center 



Pyrotechnically Actuated Systems 
Database and Catalog 

Second NASA Aerospace Pyrotechnic Systems Workshop 



February 8-9, 1994 

Sandia National Laboratories 

Albuquerque, NM 



NASA Lewis Research Center 
Cleveland, OH 



Prepared by: 
Analex Corporation 







e bvSn e engineering directorate 




Presentation Agenda 



Purpose of Database/Catalog 
Database Ground Rules 
Format for Database 
Schedule 
Database/Catalog availability 



NASA 



Lewis Research Center 




L EXCELLENCE __ ^^ 

,bv desk* ENGINEERING DIRECTORATE 



Lewis Research Center 



Purpose of Database/Catalog 



Pyrotechnically Actuated System s Database 

The purpose of the Database is to store all pertinent design, test and certification data for all 
existing aerospace pyro devices into a standardized Database accessible to all NASA/DOD/ 
DOE agencies. 



A pplications Catalog 

The purpose of the Applications Catalog is to identify and provide a quick reference for the 
pyrotechnic devices available, including basic performance and environmental parameters. 



I 



- 332 - 




EXCELLENCE 

Da " M ENGINEERING DIRECTORATE 




NASA 



Lewt» R— ewch Center 



Database and Catalog Ground Rules 



Develop database on the Macintosh computer system using OMNIS 7 software. 

Include current and past (non-obsolete) pyro devices used on launch vehicles, spacecraft, 
and support systems. Compile information from all NASA/DOD/DOE Centers. 

Include pertinent design and specification data. 

Include sketches for each device and system. 

Provide cross reference indexes in Catalog. 

Catalog to be extracted from the database. 

Provide for updating capability. 



Format 

Example I TITLE: Detonator - NASA Standard 

AGENCY/CENTER: NASA Johnson Space Center (JSC) 
PHYSICAL DATA: 



-—♦810 HEX 



N5D BODY 

.5B6 - .596 




,995 -1.000 



.39Z - .403 



-LEAD A2I0C 
-RDX 



1 



— 0.277 - .2BO 



T 



-DISCS 
-9/16 - 1BUNF-3A 
.793 - .808 



NASA STANDARD DETONATOR (NSD) SCHEMATIC 



CONTRACTOR: n/a 

SUBCONTRACTOR : HI Shear Tech. Corp., Explosive Technology Co., and 

Universal Propulsion Co. 

DEVICE IDENTIFICATION NUMBER: NASA SEB26100094 

PURPOSE : To provide a high leveled detonating Shockwave for 

initiating an explosive train or separating frangible devices. 



- 333 - 



Format (Cont.) 



^jiEfrTHp]l ^ t PREVIOUS USAGE: Apollo, Skylab, Apollo-Soyuz, and Space Shuttle. 

OPKRATIONAL DESCRIPTION i The NSD id the standard detonator for the 
Space Shuttle and is provided as GFE to all shuttle users by the 
Johnson Space Center. The NSD consists of the NASA Standard 
Initiator {NSD threaded into an A-286 stainless steel body 
containing a column of Lead Azide progressing into a column of RDX. 
When the NSI is fired with the Pyrotechnic Initiator Controller 
(PIC) 38 vcs capacitor (6B0 microfarads) discharge, the NSD produces 
a . 040 inch minimum dent into a mild steel block at ambient 
temperature. 
ENERGY SOURCE t 
TYPE OF INITIATION: NSI. 

CHARGE MATERIAL: Dextrinated Lead Azide (376 mg) and RDX (400 mg) . 
ELECTRICAL CHARACTERISTICS: n/a. 
OPERATING TEMPERATURE/PRESSURE: 
TEMPERATURE RANGE: Low -420°F, High +200°F. 
PRESSURE: n/a. 
DYNAMICS : 

SHOCK: 30g, 11 msec sawtooth. 

VIBRATION: Random <-65°F to +200°F) at 2000cps. 
QUALIFICATION: 

DOCUMENTATION: SKD-26100097 Design & Performance Spec, Qualification 
Documentation provided by each contractor and on file at JSC. 
SERVICE LIFE: 

SHELF: 4 years minimum from Lot Acceptance test date, 10 years 

maximum based upon successful passing Age Life Testing per 

NSTS 08060. 

OPERATIONAL: See Shelf Life above. 
ADDITIONAL REFERENCES: n/a. 
ADDITIONAL COMM TCNT' I ? « DOT Class-C explosive. 
SPECIAL FEATURES: n/a. 



Schedule for Pyrotechnically 


Actuated Svstems 


Database 


and Reference 


Cataloa 








Activity Name 


1993 


1994 


1995 


J ! A 


s 


O 


N 


D 


J 


F j M | A | M | J 


j 


A 


S J O ! N | D ' J 


F | M | A j M | J 


j 


A 


s 


o 


Identify all devices 










i ' j 


1 






i ' 

" f r ' 


__;_. 


'"i~- 




-.. 






Collect all data 


I 












i | 


j 






















! i ' \ 


Compile data into catalog 


^^^^^_ 












! ■ i I 

i I I 






' \ ■ 








Distribute first draft of 
catalog 
















I 
I 


! i 

i i 










i 
i 


i I : ; 


! 






Edit and revise catalog 
















! 


i 






r t 


I ! ; I i 
















1 . ! 1 






_L . 


Steering Committee approval 
of Catalog 
















i I i 
A! i ! ; 

I : i i 




— ■ 


! ; i 










Publish 1994 issue of 
NASA/DOD/DOE Catalog 


I 

! 












! i I : 






_....:_ ; _._. 


i 


■-- 










I ! i 








__t__^__;_ ,___ 


Enter information into 
database 


J 










I ! ! I 














^^^ 












! l i I 










I : s 


Distribute first draft of 
database 
















I i j I 

i i 






t i 




| ! 










Edit and revise database 




















| 








mmm 

*-':-] 

i 






( 

i i 












Steering Committee approval 
of Database 




















-- 


— 




1 


A 


I ! 

i 










Publish 1995 issue of 
NASA/DOD/DOE Database 


| 












' i 


J • i 




r ; 








i 


■ 






i ; I 








Maintain database and catalog 
thru 1995 


i 












._j — 

i 
i 






_ r " , _ 


I ! I 












1 


; 




.-._ 




i 










Steering Committee Meetings 
and status report 


A 














i 

a; 


i 




A 


1 
i 




A 


! i 

I : i 




A 






















; 












: I - 






























t 


ri 








__ 




— 


! 




.... 




— . 


— 


— 
























i 


\ i 






























J. __ 


...j..., — 












— 


i 






■ 




















Steering Committee approval milestone 










i 












rev. 1/26/94 
















i 


i 










i 


i 






j 













- 33H - 




EXCELLENCE 

vam ENGINEERING DIRECTORATE 




NASA 



Uwfc R—eareh Center 



Database and Catalog Availability 



♦ Database and Catalog will be released as government issue. 

♦ Catalog will be available in October 1994. 

♦ Database will be available in October 1995. 



- 335 - 



Page intentionally left blank 



FIRE, AS I HAVE SEEN IT 

Dick Stresau 

Stresau Laboratory, Inc. 

Spooner, WE 54801 

ABSTRACT 



Fire (and much else) is described as I have seen it (in the sense that "to see" is "to understand") from a 
succession of perspectives (which seems, like that of most moderns, to parallel the sequence of 
perspectives had been assumed by scholars in the past) since childhood. Origin and originators of some 
of these perspectives (e.g., that of the "Bohr atom" and the Chapman-Jouguet theory of detonation) are 
mentioned in passing. For the most part, I believe my views to be quite similar to those of others who 
have concerned themselves with fire in its various forms. However, data, which may not have considered 
by others, have led me to see explosion and the shock to detonation transition from apparently new 
perspectives, which are discussed. 



INTRODUCTION 

Like, I suppose, most people I first saw fire from an 
"on scene, eyewitness" perspective, Explanations, by 
adults, of fire and other things we sensed (saw, heard, 
felt, tasted, or smelled) helped me to see things in the 
sense that "to see" is "to understand", from a 
"common sense" perspective. The explanations, 
however, were in words which may not have been 
understood as they were intended. As punsters often 
remind us, each of many words and phrases of 
English, and other living languages, has a number of 
meanings. For example, the 1967 edition of the 
"Random House Dictionary of the English Language" 
(1) gives fifty-four definitions of "fire". 

PYROTECHNICS 

"Pyrotechnics" the subject of this workshop, is given 
as a synonym of "pyrotechny" - "The science of the 
management of fire and its application to various 
operations" (2). So defined, it is among the oldest of 
technologies (2V£). The survival of the human race 
(adapted as it was and is, to the climate of equatorial 
Africa) in the "temperate" zones, during the "ice ages" 
depended upon the establishment of (indoor) 
environments similar to that to which it is adapted, for 
which pyrotechny, "The management of fire" is 
essential. Its most prevalent application is still to this 
purpose. More fires are used to heat buildings than for 
any other purpose. It is an essential part of such other 
prehistoric technologies as pottery, glass making, 



metal smelting, casting and forging, as well as the arts 
of cooking, baking, etc. Remains of fireplaces are 
accepted by archaeologists, as evidence of the presence 
of early man at a site. It could be said that 
pyrotechnics, as defined above, played an essential 
role in "the ascent of the mind" (3) and the rise of 
civilization. Pyrotechny or pyrotechnics (also referred 
to, by some who practice it, as "combustion engineer- 
ing") is also part of modern practices of these ancient 
technologies and arts and of currently practiced 
technologies and arts including those of "firemen" 
(members of both fire departments and steam locomo- 
tive crews), heat-power, automotive, and jet aircraft 
engineers, (torch) welders, and heating contractors. 

As defined in most dictionaries and to most people 
who use the term, "pyrotechnics" are "fireworks", 
especially those used in public parks on Fourth of July 
evenings, Of these, the most spectacular are the 
skyrockets, which explode at their apogees in luminous 
"sprays". Such displays were made, in China, several 
hundred years B.C.(4). 

"The rockets' red glare,-" of our national anthem was 
that of military rockets, used by the British in the 1814 
bombardment of Fort McHenry, which, by the way, 
utilized and verified the capability of rockets of 
carrying substantial payloads. 

In the late 1920s and early 1930s, science fiction often 
about outer space travel, was popular (notably, with 
preteen and teen aged boys, which included me at that 



337 



time). We read that this interest had been shared, since 
their boyhood, be some scientists, including Goddard, 
in U.S.A., and Ley and von Braun, in Germany, all of 
whom made and tested rockets. 

During World War II, when most combatants 
developed rocket propelled weapons (including the 
U.S. "bazooka" ammunition), von Braun led the group 
at Peenemunde, where the German ballistic missiles, 
including the V-2, were developed. After the war, 
many of the group were recruited by D.O.D. agencies. 
Von Braun came to Redstone Arsenal to participate in 
the U. S. Army's guided missile program, and, when 
interest in "outer space" was intensified by the success 
of the Russian "Sputnik" he realized part of his 
boyhood dream as the leader of the group that put 
"Explorer I" into orbit (5). 

The events mentioned, parts of the sequence which led 
to the establishment of NASA, have led me to 
speculate as to whether von Braun or any of his 
associates or their successors in the space effort 
thought of the design and development of space vehicle 
propulsion systems as applications of "pyrotechny" or 
"pyrotechnics". They do refer to explosives and 
propellants (which Picatinny Arsenal and the American 
Defense Preparedness Association include among 
"energetic materials") as "pyrotechnics". 

LANGUAGES AND PERSPECTIVES 

As the years passed , having participated in 
conversations, committee meetings, workshops, 
seminars, symposia, etc., I have come to recognize 
that each art, profession, specialty, and often working 
group or family, communicates in its own language, 
each, in U.K., Canada, U.S., etc, a variant of 
English, using many of the same words often with 
different meanings. In recognition of the possibility of 
having been misunderstood, explanations are apt to 
conclude with, "See what I mean?, and include efforts 
to illustrate the meanings of the words by means of 
metaphors and models, such as working and scale 
models, sketches, "layout" and "detail" drawings, 
graphs, chemical formulas and equations, mathematical 
equations and sets of them, and, in recent years, 
computer manipulated numerical models, each of 
which shows the subject, from a different perspective. 
"Model" as used here indicates "a description or 
analogy used to help visualize something (as an atom) 
which cannot" (literally) "be seen" (2). It can serve 
this purpose if the analogy is to something which can 
be seen (literally or in the sense that "to see" is "to 



understand"). What some refer to as "models" are 
referred to as "theories" by others. 

Each of us sees things from a constantly changing 
perspective. In each encounter, and when we are 
reading, or writing, we try to consider each subject 
from the perspective of the speaker, writer, listener, or 
reader. Pursuant to this effort, each of us has assumed 
a succession of perspectives, those of parents, 
playmates, teachers, professors, lecturers, bosses, 
commentators, the authors of books and articles we 
have read and those of the people whose views are 
conveyed. Thus we have viewed our vicinities, the 
world, the universe, and much within them from a 
variety of perspectives, including those which can be 
categorized as "eyewitness", "common sense", 
"reasonable", "intuitive" , "practical" , "rational" , 
"theoretical" , references to those credited with 
proposing them, as "Aristotelian" , "Newtonian" , 
"Cartesian", etc., and those of several trades, sports, 
sciences, etc. 

Consideration of any subject begins with the choice of 
a perspective from which details which we wish to 
consider are visible and. others are obscured. Such 
choices are called "simplifying assumptions (or 
approximations)", in which numbers which seem too 
small to consider are dropped as "infinitesimals of the 
second order" and numbers too large to think about 
are equated to infinity (6). Sometimes such 
simplifying assumptions (or approximations) must be 
reassessed on the basis of more recent experience or 
data, which result in the consideration of a subject 
from another perspective. Such changes of perspective 
have been essential to the advance of science, and to 
the education of each individual. 

To reiterate, the perspective from which each of us 
sees things is and has been constantly changing. The 
"steam" from a teakettle and the melting of ice showed 
us that water exists in more than one state. (Other 
observations and experiences showed us that other 
substances also freeze, melt, boil and condense.). 
Steam emerging from the teakettle spout was invisible 
as air, (or, at least transparent), forming a visible 
cloud a few inches from the end of the spout. 
Someone may have explained that steam is a gas, like 
air (and thus, invisible) which, mixing with the cooler 
air, condensed to liquid water, in droplets too small to 
settle out, which we saw as a cloud. If they thought 
we could understand, they may have gone on to say 
that such suspensions of droplets of liquids or particles 
of solids, which are too small to settle out of fluids 
(liquids or gases) in which they are insoluble, are 



338 



called "colloidal suspensions" and that, when the fluid 
is air, they are called "aerosols" and are the "stuff" of 
clouds, fog, and mist, and (of other compositions) of 
smoke and smog 

FIRE, FROM "KID'S" PERSPECTIVES 

The earliest impressions of fire, which I can 
remember, were the sights of the yellow flames of 
candles on a birthday cake and of trash and wood 
fires. Fire had been the source of most artificial light 
until a bit over a century ago. As we (including the 
earliest human observers) saw the flames spread over 
the surface of the fuel, it was apparent that the light, 
which seemed to be the essential property of fire, is 
"catching", like a cold or the flu. This impression is 
perpetuated in the usage of "light" for "ignite" and 
"catch fire" for what the fuel does when lit. 

Our perspective changed with the passage of time and 
the accumulation of experience, and we came to see 
fire, from a more "practical" perspective, as a source 
of heat, and it became clear that firelight is a mani- 
festation of the heat of combustion. Some of us had 
noted that "firelight" is similar, in color to the light 
emitted by glowing coals or metal or ceramic which 
was a bit hotter than "red hot". With a little thought, 
it became apparent that black smoke, is air borne soot 
which, when hot enough, emits black body radiation, 
so the yellow flames of candles, oil lamps, and wood 
and trash fires can be seen as "yellow hot black 
smoke". Flames of other colors are effects of atomic 
and molecular emission (whereby elements and 
compounds are identified in spectroscopic analysis) 
which occur at elevated temperatures. Although the 
usage of "light" for ignite persisted in our 
conversation, we came to recognize that ignition 
occurred when a fuel element was "hot enough". By 
this time we had learned to think of heat in the 
quantitative terms of temperature and it became 
generally accepted that the temperature at which each 
fuel started and continued to burn was its "kindling 
point" or "ignition temperature". The propagation of 
fire (combustion) is seen by many as the progressive 
heating of unburned fuel to its "kindling point" by the 
heat of combustion of the burning fuel. (The title 
"Fahrenheit 451" (614) of the novel by Ray Bradbury, 
and the movie made from it, is derived from the 
supposition that 451°F is the "kindling point" of 
paper). However, the episode described below has left 
me skeptical of this view. 

When five or six years old, at a friend's invitation, I 
joined him in watching his father paint their dining 



room wall. He (the father) having painted the wall a 
light beige, let it dry, was applying a darker brown 
paint, a few square feet at a time, and, while each 
patch was "fresh", "blotting" it with crumpled news- 
paper to expose the lighter paint in a "stippled" pattern. 
He tossed the paint soaked clumps of paper into a 
bushel basket. After about an hour, the pile of paint 
soaked paper in the basket began to smoke and the 
man grabbed the basket and ran out in the yard before 
it burst into flame. He managed to drop it so quickly 
that the only damage to him was some slightly singed 
hair. A few years later, when a fireman, visiting our 
school to talk about fire and its prevention, warned of 
the danger of "spontaneous combustion", I knew, from 
on scene, eyewitness observation, what he was talking 
about, from the "practical" perspective of the fire 
fighter, but didn't see why "spontaneous combustion" 
could happen if the "ignition temperature" of the paper 
was as high as it seemed to be. 

At the time of the above episode, our family lived in 
a suburb of Milwaukee. Sunday drives (in the "Model 
T") often carried us to deserted stretches of Lake 
Michigan beach, on which there were, often 
"fireplaces", made by arranging rocks in a circle. My 
Dad often built fires of driftwood , which was usually 
available. If suitable vegetation grew nearby, he'd 
sharpen "green" branches or "shoots" to make "spits" 
for toasting marshmallows or broiling hot dogs. As we 
sat around the fire I saw it as the focal point of the 
family's togetherness". 

A few years later, I went with my parents on an 
automobile trip around Lake Michigan. Most 
memories of the trip, particularly that of the ferry boat 
crossing of the Straits of Mackinac, are pleasant, but 
one, less pleasant but more vivid, is that of mile after 
mile of "burnt over country" left by the "Peshtigo 
fire", which, fifty years earlier (on the date, Oct. 20, 
1871, of the more notorious though less disastrous 
Chicago fire.) had devastated 1 ,280,000 acres, 
including three-quarters of the shores of Green Bay 
(7). Even after fifty years, the effects of the fire were 
quite apparent. I suppose that my parents thought that 
I had been sufficiently persuaded of the importance of 
keeping fire under control to trust me, as they did, 
with the responsibility of burning trash (in a woven 
wire trash burner) and leaves (in the gutter, as was the 
practice in our oak shaded neighborhood). 

My household duties, as subteen, besides trash 
burning, raking and burning leaves, and lawn mowing, 
included tending the coal furnace, which involved 
shaking out ashes, adjusting the damper when neces- 



339 



sary, and shoveling coal, in the course of which I had 
plenty of opportunity to observe the fire, which was 
mostly glowing coals, with a few flickering flames. 
When the fuel was coke, there were no flames nor 
smoke. 

After a few years, as a Boy Scout, I became involved 
in a discussion of the concept of "kindling points" or 
"ignition points". In the course of the discussion, I 
recalled die "spontaneous combustion" episode I'd 
seen. By that time, of course, I'd forgotten, if I ever 
knew, the temperature on the day when I'd witnessed 
spontaneous combustion but guessed that it must have 
been between 60°F and 90°F and wondered, out loud, 
whether paint soaked newspaper had a kindling point 
in that range. The answer was that it didn't but that 
"self heating" had raised the temperature of the stuff 
in the bushel basket to its kindling point. It would be 
more years before I could sort out the distinctions 
between "self heating", "burning", "fire", and 
"combustion". (Perhaps I haven't yet, but con- 
sideration from various perspectives has helped me to 
understand those who use these words.). 

Based on our earliest impressions of such matters, the 
aphorism that, "Where there's smoke, there's fire.", 
seemed to be a "matter of common sense". However, 
when we saw smoke, but no flames, coming from an 
overloaded extension cord or from toasting bread or 
frying bacon that was getting black and said to be 
"burning", with no flame evident, and that, as fire 
spread across burning wood, smoke often appeared 
ahead of the flame, we pondered the questions as to 
what was meant by such words as "smoke" and "fire" 
and, specifically, what is the composition of smoke 
(including, the liquid "hickory smoke" and "mesquite 
smoke" in bottles on grocery store shelves). We'd find 
answers in considering these questions from an 
"intermolecular" perspective of fire and fuels. 

We may have noticed that, if metal ceramic or 
anything else that can stand the heat, is heated 
sufficiently, it gets "red hot" and, if heated more, the 
red brightens to orange, yellow, and, with further 
heating, the object becomes "white hot". Although I 
don't remember hearing (he phrase, "yellow hot", it 
seems an appropriate designation of a condition 
between "red hot" and "white hot". We may have 
heard or read that such glow is called "black body 
radiation", although a red hot poker doesn't look 
black. 

Some of us noticed that, the flames of candles and 
kerosene or alcohol lamps are yellow, those of a gas 



stove are blue (if the burner is "properly" adjusted, 
but yellow if the air intake is restricted to make a 

"rich" mixture. 

We learned, when quite young, that, although candles 
and matches could be "blown out", fire, in general, 
was energized by blowing or "fanning" ("red hot" 
glowing coals brightened and became "yellow hot"), 
controlled by "damping the draft" and "smothered" by 
depriving it of air. It is apparent that the foregoing 
was known by prehistoric humans. Smelters, 
foundries, and forges in archeological sites had flues, 
dampers, and other means of forcing and limiting the 
supply of air to fires. 

Those of us who are old enough to have had spending 
money in the 1920s (before the passage of "safe and 
sane Fourth" ordinances) celebrated the Fourth of July 
with firecrackers and other fireworks (some, 
apparently, still do), in the course of which, we 
learned that gunpowder and other pyrotechnics burn 
without air (and, in fact, burn faster when confined - 
"smothered"), which had been known by some 
alchemists as long ago as the ninth century (4). 

"Educational toys" included chemistry sets, which 
provided amusement, seeing the color changes and 
foaming resulting from chemical reactions, and, 
perhaps, the beginning of a "chemical" perspective. 
While, since "playing with fire" was discouraged, in- 
gredients of gunpowder and similar mixtures were not 
included in chemistry sets, some of us learned (by 
reading) what they were and found ways to get some, 
and set out to make our own fireworks. I joined a 
fourth grade classmate in the "Universal Research 
Laboratories" as he called his basement in the 
preparation of some black gunpowder which we loaded 
into a home made skyrocket (which flew straight up to 
about six feet from the ground before it lost stability 
and tumbled). In the course of these efforts, we 
viewed chemistry from the "practical" perspective 
from which it seemed that what had worked for others 
should work for us (a perspective shared with the 
Middle Ages alchemists who had practiced pyrotechny 
since antiquity). 

A "CHEMICAL" PERSPECTIVE 

When, in high school, we were shown a "chemical" 
perspective, we saw that (as Lavoisier had shown in 
1779 (5)) the fire we had seen (from an "eyewitness" 
perspective) was "oxidation" (combining with the 
oxygen of air) and that gunpowder and other pyro- 
technics burn without air because they are mixtures of 



340 



fuels with nitrates, chlorates, or other compounds 
which decompose when heated releasing oxygen which 
reacts with the fuels. All of which is expressed in 
stoichiometric equations, such as: 

2H 2 +0 2 -»2H 2 (1) 

for the reaction of hydrogen and oxygen, and: 

2KN0 3 +C + S-» C0 2 +S0 2 +2KO (2) 

for the burning of black powder of stoichiometric 
composition. 

In stoichiometric equations, as equations (1) and (2), 
the symbols (H, O, C, K, N, and S) stand for 
elements (hydrogen, oxygen, carbon, potassium, 
nitrogen, and sulfur) and the formulas, which are 
essentially inventories of the proportions of the 
elements of which each compound (water (H 2 0), 
carbon dioxide (CO^, potassium nitrate (KN0 3 )) are 
referred to as "empirical" formulas, since they are 
based on empirical (experimental) data as are valences 
(upon which predictions are made, of formulas of 
compounds which have yet to be made) assigned each 
element. The small integers, which express these 
proportions, suggested to Proust (in 1799) and 
corroborate the "law of definite proportions", which 
suggested (in turn, in 1801, to Dalton) the basis of 
modern atomic and molecular theories. Atomic 
theories were proposed, some hundreds of years 
B.C., by Pythagoras, Democritus, and Lucretius (5,8). 
The modern theories are based on and supported by 
empirical data. Some see atoms as portrayed by 
models. The structural formulas, which are used to 
represent organic compounds are models (diagrams) of 
their molecules. Some organic chemists, before trying 
to make a compound, try to build a scale model of its 
molecule, in which the atoms are represented by 
plastic spheres of various sizes and colors. As I 
understand it, the colors are only for identification of 
the elements represented but the sizes are scaled 
(typically 2 or 3 centimeters per angstrom) from the 
effective sizes of the atoms represented. Similar 
models (typically, styrofoam balls, joined by tooth- 
picks (Figure 1) are used for educational purposes. 



Figure 1. Educational Models of Molecules (enlaced, 
from Edmund Scientific Co. catalog) 



The perspective induced by such models can be 
referred to as an "intermolecular" perspective. Fire 
can be seen from this perspective, in the sense that "to 
see" is "to understand", only with reference to other 
perspectives, views from some of which have been 
mentioned in the foregoing discussion. Consideration 
from "practical" , "empirical" , "scholarly" , and 
"theoretical" (including laws of gravity, magnetism, 
fluid mechanics, thermodynamics, heat transfer, and 
reaction kinetics is necessary to "see" fire clearly. 

"PRACTICAL" PERSPECTIVES 

Based on some of our earliest experiences, such as 
falling down and dropping things, and observations, 
such as those of falling objects and the flight of balls, 
led us to accept the aphorism that "what goes up, mist 
come- down.", which I've heard cited (on television) 
as "the law of gravity". Rubber band (referred to, by 
some, as "elastic bands" showed us that some things 
are elastic, that is,when deformed, they tended to 
recover their pervious shape. These and similar ex- 
periences and observations gave us,when we were very 
young, a practical, through rudimentary perspective of 
gravity and mechanics. 

Most of us played with magnets, usually horseshoe 
shaped steel items, painted red, except at the ends, 
which picked up nails, pins, etc, to which they came 
close. If we had two magnets, we found, after a few 
tries, that they attracted or repelled one another, 
depending upon which end of one was close to which 
end of the other. Someone older, probably told us that 
the ends were called "poles", one "north" and the 
other "south" and that opposite poles attract, and like 
poles repel one another. They may have gone on to 
say that the earth is a giant magnet and the needle of 
a compass a tiny one, which aligns itself with the 
earth's magnetic field. 

Our earliest impressions of electricity were related to 
its practical applications. Lights could be turned off 
and on from across a room or upstairs from downstairs 
or vice versa, by "closing" or "opening" a switch. 
Vacuum cleaners, electric fans, and washing machines 
ran if "plugged in" and "turned on". When we first 
learned of such possibilities, electricity .seemed akin 
to magic. Play, with electric trains, the small motors 
which came with Erector sets, and the lighting fixtures 
associated with them improved our "practical" 
perspective of electricity. The electricity involved in 
such play came from transformers, which were "plug- 
ged in". We learned, quite early that a direct connec- 
tion of the - terminals of a transformer made it hum 



341 



quite loudly and get warm. We were told to "break" 
the "short circuit" before it "burned out" the trans- 
former. 

Some of us learned that batteries could be used instead 
of transformers to run electric trains, etc. We all 
became aware of batteries as parts of flashlights and/or 
battery operated toys. We found that batteries differed 
from transformers in three ways; 

1. They have "positive" and "negative" terminals, 
marked " + " and"-". 

2. They discharge as they are used and in a short 
time, when "shorted". 

3. They don't hum, even when "shorted" (but do 
warm). 

We were told that the reason for these differences was 
that the "electrical current" from a battery is "direct 
current" - "D.C.", which flows, without variation, in 
the same single direction, while the current from a 
transformer, and "house current" are "alternating 
current" - "A.C." which flows, alternately, in opposite 
directions. Since the direction changes and changes 
back again sixty times per second, it is called "sixty 
cycle", or, in recent years, "sixty hertz", current. The 
reason most "house current" is A.C. is that its voltage 
can be changed by means of transformers. Most of us 
had learned, when quite young, that one could get a 
"shock" from 110 volt "house current", but not from 
the 5 to 10 volt output of a transformer. Later we 
learned that the compelling motive for the use of A.C. 
by utilities was the greater efficiency of transmission 
at thousands of volts, which would be unsafe as "house 
current", so the high voltage of the transmission lines 
is transformed to 1 10 volts by a transformer near each 
point of use. 

With the passage of time, we acquired some 
"practical" perspectives of household and automotive 
electrical systems and the magneto powered ignition 
systems of outboard motors, chain saws, and lawn- 
mowers, all of which are powered by "internal 
combustion engines" in which a fuel-air mixture is 
ignited by "spark plugs" which emit sparks when 
actuated by electrical pulses from their "ignition 
systems". To many "spark" became synonymous with 
"ignition" (a matter to be discussed when we get to 
"ignition"). 

In the late 1920s, as a result of the interaction of 
advancing technology, patent law, and business 
competition the best radios were home made, so lots 
of people made their own (which were battery 



powered). By the early *s, the "art" had advanced and 
"store bought", "plug in" radios replaced the home 
made battery sets the parts of which became available 
to teenagers for basement experimentation, from which 
we gained "practical" perspectives of electronics. One 
misconception, which was corrected, by the view from 
the "electronic" perspective was that electrical current 
flows from positive to negative. We became aware 
that electrical current is the movement of negatively 
charged electrons, from negative to positive. We had 
previously learned, from demonstrations of 
electrostatic effects (with combs and bits of paper), 
that like charges (as like magnetic poles) repel one 
another, and opposite charges attract. We were to 
learn, later, that these principles apply to chemical 
reactions, including fire as well as electrolysis. 

"Practical" perspectives are acquired, along with 
"know how", skills, or "arts", by imitation, 
instruction, and experience. Experience is gained by 
"trial and error" (referred to, by old time machinists 
as "cut and try" and, by those who would dignify it, 
as "Edisonian research"), followed by practice of 
techniques which were found to be successful. 
"Practical" perspectives are those from which we 
(including everyone who has ever lived) consider the 
application of things and materials to immediate or 
anticipated purposes. The "practical" motorist is 
aware that the high compression engine of a 
"Corvette" will run best on "high octane" gas (which 
may cause a "Model T" to overheat) but may not have 
considered, from a "scientific" perspective, the reasons 
for a high compression engine, nor why "high octane" 
gasoline does what it does. My impressions of such 
matters go back to the late twenties, when "Ethyl" and 
"Benzol" pumps began to appear at gasoline stations 
and we heard that it was needed for the newer cars 
(like the recently introduced Chrysler) with "high 
compression" engines (7 to 1 was considered "high") 
which would "knock" on "regular" gas". I saw and 
heard it tried and the "knocks sounded as if something 
was trying, with a hammer, to beat its way out of the 
engine. I was told that knock was the result of the 
"regular" "burning too fast" at high pressures and that 
the burning could be slowed by adding tetraethyl lead 
or benzine to make "Ethyls" or "Benzol". Still later, 
I learned that gasoline was rated for its resistance to 
"knocking" by means of the "Research Method", 
which involves the use of an internal combustion 
engine of adjustable compression ratio, which has been 
calibrated using mixtures of isooctane (100 octane) and 
normal heptane (zero octane) (10), I have never had 
occasion for further consideration of such tests. I 
have, somehow, gained the impression that "knocking" 



342 



of an internal combustion engine has been ascribed to 
"detonation" of the fuel-air mixture. In this respect, 
I don't know don't know which of the several 
meanings of "detonation" was intended, but would 
guess the definition of Chapman(12) and Jouguet(13), 
which is reviewed in the later section hereof headed - 
EXPLOSION AND DETONATION. 

Also, in the late 1920s (I was in my early 'teens), I 
was on a seagull banding expedition (for the National 
Bird Survey) in northern Lake Michigan aboard a 
Diesel powered Coast Guard tug (similar to the fishing 
tugs which were common in the upper lakes at that 
time). The Diesel engine, as I recall, was about six 
feet high and ten feet long. Preparatory to starting it, 
the engineer lit a blow torch over each of the four 
cylinders. When they were hot enough, he ran the 
engine as a compressed air motor to "crank it", after 
which he quickly reset hand operated valves and the 
engine ran as a Diesel engine. I was told that, in a 
Diesel engine, the fuel (kerosene - or "coal oil", as 
some called it then) was ignited by "compression 
ignition" rather than a spark". My dad and the 
organizer of the expedition (to whom bird banding was 
a combination hobby and public service activity) were 
engineers by profession. One of them explained the 
"Diesel cycle" to me, about (as remembered after six- 
ty-some years) as follows; 

Air, which had been drawn into the cylinder 
in the intake stroke, is compressed "adia- 
batically" (my introduction to this word, and 
to the subject of thermodynamics) which 
raises its temperature above the ignition 
temperature of the fuel, which burns as it is 
injected, (the engine can't "knock" because 
the fuel can't burn any faster than it is 
injected into the hot air). The heat of 
combustion of the fuel raises the temperature 
and hence the pressure of the air and product 
gases, which expand adiabatically, imparting 
more mechanical energy to the system than 
was used in the compression stroke. 

The above explanation, in combination with the 
rationale that Diesel engines, which have higher 
compression ratios than gasoline engines, should be 
more powerful and efficient, besides which, they used 
cheaper fuel. All of which persuaded me that auto- 
mobiles should have Diesel engines. Based on this 
conviction, I enrolled, a few years later (in 1934),- at 
M.I.T. in Course IXB, "General Engineering" (in 
which each student could choose courses appropriate 
for his intended specialty) to specialize in automotive 



Diesel engines. After a year or so. recognizing that I 
was taking the same courses as those in Course II 
"Mechanical Engineering" whose schedules had been 
prearranged, so I switched course to avoid the hassle 
of trying to arrange my own, while others in my 
classes were doing the same, These moves were 
motivated by the recognition that practical objectives 
are most attained by those who understand the prin- 
ciples involved. 

This transition, of my perspective, from "practical" to 
the "scholarly" is one of many I, and, seemingly many 
others, have made since humans became human. 

"SCHOLARLY" PERSPECTIVES 

As the word implies, a "scholarly" perspective was 
gained in school. That acquired in the lower grades is 
scholarly" in the sense that, like that of the Medieval 
"Scholastics", it presented the perspective from which 
the world was seen a few millennia ago, when the 
classics, which, often, explained phenomena with 
reference to such models as anthropomorphic animals 
(e.g. the tortoise and the hare) and objects (the 
mountain which talked with the squirrel), supernatural 
entities such as, fairies, brownies, trolls, gods, and the 
heroes of Greek, Norse, etc. mythology, were written. 
From a modern perspective, it is, sometimes, hard to 
tell where the line was drawn between the 
metaphorical and the literal. Even in modern 
discourse, the location of this line is sometimes 
indefinite. As we progressed in school, the, 
"scholarly" perspective melded into "classical", 
"historical", "mathematical" and "scientific" 
perspectives, which were narrowed to those of specific 
n subjects" or, "academic disciplines", such arithmetic, 
science, algebra, geometry, chemistry, and physics, 
and further, to trigonometry, calculus, organic and 
physical chemistry, applied mechanics, 
thermodynamics, electrostatics, and vector analysis. As 
a result we saw fire and other phenomena and things 
from a succession of perspectives, between which we 
shifted, often after a few seconds. 

"Classical" perspectives included those mentioned 
above based on mythology and those of Greek 
philosophers, including Plato, Euclid, Pythagoras, and 
Aristotle. Euclid's geometry in which the subject is 
considered from a "logical" perspective, is, in English, 
still taught. Aristotle viewed physical science from a 
"logical" perspective in which phenomena are 
explained in terms of relationships derived from "first 
principles", which, like the axioms of geometry were 
considered (on the basis of what some modern 



343 



scientists view as "intuition" when they refer to a 
phenomenon as "counter-intuitive") to be "self evident 
truths". While the axioms of geometry have stood the 
test of time, some of Aristotle's "first principles", such 
as that "heavy objects fall more rapidly than light 
ones" were discredited by empirical data. 
Consideration of phenomena from the "empirical" 
perspective, followed by views from "graphical", 
" analytical " , and " theoretical " perspectives has 
advanced science since the renaissance. Empirical 
data are quantitative data, the product of measurements 
of physical quantities, including time, dimensions, 
position, force, mass, etc, and functions thereof, such 
as velocity, acceleration, pressure, energy, and power. 
Some such measurements (for example, those of 
astronomers, microscopists, etc.) are made, using 
instruments, of natural phenomena which are beyond 
the control of the observers. Others are made in the 
course of experiments, including the establishment of 
preconditions and determination of results. By 
"plotting" such data a "graphical" perspective is 
gained, the view from which may suggest 
relationships, which can be expressed in algebraic 
equations, manipulation of which can yield an "analyt- 
ical" perspective. Consideration of an object, material, 
system, or phenomenon from "empirical", "graphical" 
and/or "analytical" perspectives may lead to the 
conception of a model or theory, usually, at first 
"heuristic" but, for purposes of discussion and 
analytical verification, represented by a diagram, 
mathematical or chemical formula, equation or set of 
equations, graph or scale model, which provides 
another perspective, which can serve as the basis for 
verification, or at least support, of the theory by 
prediction of observed or experimental data. Such a 
sequence has resulted in the advance of each science. 
However, since the sciences differ with respect to the 
phenomena and quantities with which they are 
concerned, each has evolved an unique perspective 
(each of which is the result of such a sequence). 

From "hydraulic" perspectives, water and other liquids 
are seen as "incompressible fluids" which behave in 
accordance with Pascal's law, that "Pressure (force per 
unit area) exerted at any point on confined liquid is 
transmitted, undiminished, in all directions" (10). The 
"hydrostatics" perspective is that from which systems 
in which the effect of flow upon pressure is negligible, 
so that Pascal's law can be applied without reserva- 
tion. 

Systems and phenomena in which the effect of 
movement upon pressure is significant are considered 
from the "hydrodynamic" perspective, from which this 



effect is seen to be determined by Bernoulli's principle 
(which is the law of the conservation of energy stated 
in terms of pressure, density, and velocity (5,11). 

Although, for liquids, the assumption of 
incompressibility is a reasonable approximation (in 
fact, when the empirical data, upon which the 
principles of hydraulics and hydrodynamics were 
based, were obtained, the available instruments were 
sufficiently precise to determine the compressibility of 
liquids, the compressibility of gases was apparent to 
Hero in the first century (5) and must be taken into 
account in considering their behavior. The behavior of 
gases is considered from a "thermodynamic" per- 
spective in terms of the "gas laws", which relate 
pressure, specific volume, and temperature. 

Liquids and gases are seen, from "hydraulic", 
"hydrodynamic", and "thermodynamic" perspectives 
(as they are from "eyewitness", "common sense", 
"intuitive" "practical" and "empirical" perspectives) as 
continuous media (thus legitimizing the application of 
algebra, calculus, and differential equations in the 
generation of theory from empirical data. In contrast, 
from the "intermolecular" perspective, which is 
mentioned a few pages back, a gas is seen as a 
"swarm" of atoms and molecules, moving in random 
directions at random velocities, and bouncing, 
elastically, from one another ("like tiny billiard balls" 
as Mach, scornfully, put it) when they collided, as 
envisioned by Maxwell and Boltzmann, who 
considered this motion in terms of statistical 
mechanics, assuming what has become known as the 
"Boltzmann factor", exp(-E/RT), for the statistical 
distribution of kinetic energy, E, among the molecules 
and atoms of a volume of gas, at absolute temperature, 
T, where R is the "gas constant" for the mechanical 
equivalent of heat. The result of their consideration (in 
1 87 1) from this "intermolecular perspective has 
become known as "The Maxwell-Boltzmann kinetic 
theory of gases", that heat is molecular motion and 
pressure is the aggregate effect of impacts of many (if 
the order of Avogadro's number (6.026X10 23 ) times the 
"Boltzmann factor") molecules on a surface. That the 
"gas laws", which had been established in the previous 
century, on the basis of experimental data, by Boyle 
and Charles, can be derived from the kinetic theory of 
gases has validated the theory. 

As Mach's remark, alluded to in the preceding 
paragraph, implies, heat is viewed from more than one 
perspective. Some, apparently including Mach, view 
it as a fluid (as it seem to be from casual observation 
as well as carefully controlled "heat flow" experi- 



344 



ments). Even today, though heat is generally seen as 
"molecular motion", the phrase "heat flow" is 
common. Although the Maxwell-Boltzmann kinetic 
theory of gases, when generally accepted, established 
the view that heat is molecular motion, I'm not sure 
that Maxwell or Boltzmann considered the motion to 
include rotation and/or vibration. 

Watt's invention of the steam engine (in 1769) and its 
widespread application motivated the development of 
the thermodynamics of steam, in the course of which 
it became evident that, although the "gas laws" apply 
to "superheated steam", they don't to "wet steam", so 
that "steam tables" and "Mollier charts" were needed 
for quantitative prediction of operating characteristics 
of steam engines. Van der Waals considering such 
"two phase" systems from the "intermolecular" 
perspective of Maxwell and Boltzmann, assuming 
finite sizes of (and attractions between) molecules 
derived the "equation of state" ,which is known by his 
name, in 1873. 

Chemistry, considered from the "empirical" 
perspective, had suggested and corroborated the "law 
of definite proportions" which, in turn, suggested the 
atomic and molecular theories. Faraday, viewing 
electrolysis from an "empirical" perspective, 
established his "laws of electrolysis", which introduced 
the concepts of "equivalent weight" and valence, in 
1832 (5). (He also favored the proposition that electric 
current is composed of particles (something that 
Franklin had suggested nearly a century earlier 
(incorrectly assuming the particles to have what he 
designated, and is still referred to as a "positive" 
charge) and would be verified, in the 1890s, by 
Arrhenius and Thomson, who corrected Franklin's 
error) (5). 

" Equivalent weights " as measured by Faraday ' s 
methods, are ratios of atomic weights or valences. By 
1860, the lack of consensus regarding the means of 
separating atomic weight from valence had chemistry 
in a state of controversy and confusion which moti- 
vated the convening of the First International Chemical 
Congress, ins which these matters were resolved. By 
the late 1860s, atomic weights and valences of all 
elements known at the time had been determined. 
When Mendeleev tabulated the elements in order of 
atomic weights, and entered valences in the table, he 
noted a periodicity of the valences, and in 1871, he 
published the, now ubiquitous, Periodic Table of the 
Elements. 



As pointed out and illustrated herein, perhaps too 
often, each of us sees things, substances, systems, and 
phenomena from a sequence of constantly changing 
perspectives, and the sequences are individual. In my 
case, the perspectives alluded to in the foregoing were 
gained by observation, experience, study in school, 
and recreational reading. I don't remember the exact 
sequence, but it seems that before I acquired an 
"intermolecular" perspective (in fact, before the 
styrofoam used to make the models shown in Figure 
(1) was invented) I had read a book (9) which induced 
an "intra-atomic" perspective, from which I saw an 
atom as a "nucleus" of closely packed protons and 
neutrons, surrounded (as the sun is by planets) by elec- 
trons. 

AN "INTRA-ATOMIC" PERSPECTIVE 

Each electrostatically positive proton attracts an 
electrostatically negative electron, so the electrons, 
which don't fall into the nucleus (as the planets don't 
fall into the sun) because they are moving too fast. 
Thus, the electrons orbit about the nucleus as the 
planets do about the sun. The analogy of an atom to 
the solar system is flawed by the difference between 
gravity, which attracts all heavenly bodies to one 
another and the electrostatic forces, which attract 
electrons to protons but repel them from one another. 
The orbital patterns of electrons are determined by the 
interaction of these forces and the laws of motion 
established by Newton's demonstration that they can 
be invoked as the basis of the derivation of Kepler's 
(empirical) laws of planetary motion (as well as the 
principles of relativity, and wave and quantum 
mechanics postulated, formulated and demonstrated in 
the early 20th century, by Einstein, Planck, Bohr, 
Pauli, Heisenberg, and others (8)), of which my 
understanding was (and still is) too vague to include in 
the model upon which my "intra-atomic" perspective 
was based). In this model, the equilibrium positions of 
the electrons, as determined by their attraction to the 
nucleus and their repulsion of each other is attained 
when they are spaced in an orbit where the attraction 
of the electrons to the nucleus is balanced by their 
repulsion of each other (plus the centrifugal force due 
to their motion in that spherical surface). Two 
electrons can occupy such an orbit, from which 
additional electrons are excluded in accordance with 
the Pauli exclusion principle, (that only two electrons 
(with opposite "spins") can occupy any given 
"quantum state") (8). It seemed that, in the language 
of atomic physics, the term "orbit" meant the spherical 
surface (referred to, by some (5)(8) as a "shell") 
where these forces are in equilibrium such that the 



345 



attraction of the electrons to the nucleus balanced their 
repulsion of one another plus the centrifugal force due 
to their motion in this spherical surface and the effect 
of the Pauli exclusion principle etc. Electrons which 
are excluded from this inner orbit locate in 
surrounding orbits, which are larger because the net 
attractive force of the nucleus has been diminished by 
the repulsive force of the electrons in the inner orbit, 
so there is room for eight electrons to attain 
equilibrium positions in compliance with the Pauli 
exclusion principle. A third orbit has room for eight 
electrons, while the fourth and fifth orbits contain 
eighteen electrons each and the sixth has room for 
thirty-two. A seventh orbit, presumably could contain 
thirty-two, if and when elements that heavy are dis- 
covered. 

The foregoing is a description of the heuristic model 
of an atom, which I remember, after sixty years as the 
basis of the intra-atomic H perspective got from reading 
"Inside the Atom", by Langdon-Davies (9), as well as 
high school courses I had completed in chemistry, 
physics, and solid geometry. In my reconstruction of 
the model, I was helped by the copy of the book (9), 
which I got for Christmas in 1933 and still have, as 
well as more recent publications, including periodicals 
and references (5), (8), and (10), by university courses 
in chemistry, physics, mechanics, thermodynamics, 
physical chemistry, fluid mechanics, etc and from 
conversations with and lectures by scientists, including 
Gamow, Eyring, and Kistiakowski, all of which may 
have "edited" my memory of some details. It is 
apparent, from "browsing" through references (5) and 
(8) , that the model described above is an 
approximation of that which is sometimes referred to 
as "the Bohr Atom", which is the result of the work, 
guided by Niels Bohr at the Bohr Institute, by an 
international group of physicists, including Heisenberg, 
Pauli, Gamow, Fermi, and Oppenheimer who 
considered their work in progress, from perspectives 
of earlier contributors to science, including 
Pythagoras, Dalton, Avogadro, Mendeleev, Thomson, 
Maxwell, Boltzmann, Van der Waals, Rutherford, 
Planck, and Einstein, to mention a few (8). As a high 
school senior, I saw the model, as one sees a ship on 
a foggy night (only by its running lights), befogged as 
it is by relativity (which equates matter to energy) and 
wave and quantum mechanics (which consider light 
and electrons, seemingly alternately, as waves, 
particles, vibrations, and/or orbits(8)) and my present 
view is still quite misty. The model described above, 
however, has clarified, somewhat, my view of 
chemistry, electronics, and thermodynamics. 



AN "INTERATOMIC" PERSPECTIVE 

The electrostatically negative field of a "saturated" 
(filled) electron orbit or "shell", in combination with 
the Pauli principle, results in a repulsion of electrons 
or other electron "shells" which increases with 
proximity so abruptly that (for atoms of so called 
"monatomic", "noble", or "inert" gases (helium, neon, 
argon, krypton, xenon, and radon) all of whose 
electron orbits are saturated), Mach's reference to the 
molecules (which include atoms of monatomic gases) 
in accordance with the Maxwell- Boltzmann kinetic 
theory of gases as that of "tiny billiard balls" can be 
considered an accurate analogy. 

Of the hundred plus known elements, only the six inert 
gases mentioned above have saturated" outer electron 
orbits. Atoms with unsaturated outer orbits join in 
groups in which the electrons of the unsaturated outer 
shells (orbits) are shared. The most familiar grouping 
(from a "chemicals' perspective) is in molecules, 
where atoms with relatively few electrons in their 
outer shells (metals, such as sodium, which has one) 
combine with those with nearly saturated outer orbits 
(nonmetals, like halogens, including chlorine, whose 
outer orbits lack one electron each of saturation, in 
which all orbits are saturated, such as that of sodium 
chloride (NaCl table salt). 

The electrons of the unsaturated outer electron orbits 
are referred to as "valence electrons" because their 
numbers correspond with the valences of the elements 
to which they apply. 

As mentioned a few pages back, Faraday had 
introduced the concept of valence in 1832, and 
Mendeleev had published his Periodic Table in 18771. 
The "Bohr atomic model", roughly described above, 
was conceived, in part, as an explanation of the 
empirical evidence of the periodicity indicated in 
Mendeleev's table. 

The "valence electrons" of two or more atoms are 
drawn into saturated orbits by some of the forces 
whose equilibrium determines the numbers of electrons 
in the saturated orbits (or " shells " ) while the 
electrostatic equilibrium between the protons and 
electrons of each atom hold it together. Thus 
molecules are formed in which atoms are so grouped 
that they can share electrons to saturate all orbits while 
the electrostatic equilibrium of each atom is 
maintained. Such combinations of forces, which hold 
molecules together, are called molecular bonds. Figure 
1 shows models of such molecules, of which the 



346 



"billiard balls" are a less accurate analogy than they 
are of the on atomic molecules of "inert" gases in that 
they are assemblies of spheres rather than separate 
balls, so that their movement, involving significant 
fractions of their kinetic energy ("heat") includes that 
of rotation and vibration. While models, such as those 
shown in Figure 1, are usually scale models they 
cannot be viewed as "working" models, because the 
"atoms" are rigidly joined, while the real atoms of real 
molecules assume equilibrium relative positions 
(determined electrostatic, electromagnetic, and other 
forces involved in the for saturation of atomic electron 
orbits) about which they vibrate or orbit (orbiting is, 
essentially, vibration in two or three dimensions.) 
Consideration of heat as molecular motion, from this 
perspective has led to the distinction between 
"rotational", "vibrational", and "translational" heat, 
which accounts for the differences between gases with 
respect to ratios (k=C p /C v ) of specific heat at constant 
pressure (C p ) to that at constant volume (C v ). 

Molecules are groups of atoms held together by the 
combination of the quantum mechanical forces which 
establish conditions for saturation of atomic electron 
orbits end the electrostatic equilibrium which results in 
the equality, in each atom, between the number of 
positive protons in the nucleus and the total number of 
electrons which orbit about it. The repulsion of 
"saturated" orbits for additional electrons (including 
those in other "saturated" orbits) increases so sharply 
with proximity that the comparison to "tiny billiard 
balls (or golf, tennis, ping-pong, or basket balls) is an 
accurate analogy of the bounce of colliding molecules, 
atoms, or groups of atoms, including "ions" (atoms or 
groups of atoms of which all electron orbits are 
saturated). The forces which hold the electrons in 
orbit, and thus maintain saturation, combine with the 
electrostatic force which maintains the equality 
between the nuclear protons and orbiting electrons of 
each atom to result in attraction which, like gravity, is 
relatively constant close in and, varies inversely as a 
function of greater separation so that that which is 
referred to as "vibrational heat" is alternate collision 
and separation of atoms and groups of atoms, 
including "ions" (atoms with outer orbits saturated by 
the transfer of electrons, which leaves each ion with 
an electrical charge). Those with more electrons in 
orbit than protons in their nuclei have negative charges 
and those with less have positive charges. The 
attraction between the atoms and groups of atoms 
(including ions) of a molecule (in the gaseous state), 
like the gravitational field of a planet, varies so little 
"close in" that it can be viewed as constant and varies 
as an inverse function of greater distances, while their 



repulsion of one another increases with proximity, as 
sharply as that of elastic solid objects. Thus, the 
motion of those components of molecules associated 
with "vibrational heat" is more analogous to that of 
bouncing balls than to that of vibrating piano strings. 

AN "INTERMOLECULAR" PERSPECTIVE 

The specific heat of a gas is the quantity of energy 
associated with a one degree increase in the 
temperature, of a unit quantity of the gas. Considered 
from the Maxwell-Boltzmann intermolecular 
perspective, heat is the kinetic energy of molecular 
motion. At constant volume, the energy required to 
heat a given quantity of gas one degree is only the sum 
of the corresponding "translational", "vibrational", and 
"rotational" motion of the molecules, while, at 
constant pressure, the energy involved in the expansion 
is added. Since only translational heat is involved in 
expansion, where vibrational and rotational heat are 
significant fractions of specific heat, ratios of specific 
heat at constant pressure to specific heat at constant 
volume are reduced. 

From the perspective of the Maxwell-Boltzmann 
kinetic theory of gases, only the mutual repulsion of 
molecules, atoms, and ions (at very close proximity), 
(which result in effectively elastic rebound from 
collisions, like those of "tiny billiard balls") are dis- 
cernible. As mentioned, the empirically established 
gas laws of Boyle and Charles can be derived from the 
theory. However, from the "interatomic" perspective 
discussed in the previous section hereof, molecules, 
except for those of inert gases, are more complex than 
balls . A more accurate analogy would be an 
assemblages of balls as represented by the models 
shown in Figure 1, except that they are not rigidly 
connected, but vibrate about equilibrium relative posi- 
tions, since molecules, atoms, and ions are attracted to 
and repelled from one another by forces (mostly 
electrostatic and magnetic) which vary with their 
degrees of proximity, relative positions, and 
orientations. From this perspective, it is apparent that 
the kinetic theory of gases, including the "Boltzmann 
factor", applies to "vibrational" and "rotational" as 
well as "translational" heat. 

NOTES ON THERMODYNAMICS 

Carnot "founded" (5) the science of thermodynamics 
with his book, a partial title of which is "On the 
Motive Power of Fire", in which he cited empirical 
data showing that, in expansion of a gas, heat is 
transformed into "work" and conversely, in 



347 



compression, "work" is transformed into heat, a half 
century before Maxwell and Boltzmann presented their 
kinetic theory of gases (which includes the postulate 
that that which is sensed and measured as pressure of 
a gas is the integrated effect of impact of molecules 
upon the surface). If, as implied by the "tiny billiard 
ball" analogy, the molecular motion, which the kinetic 
theory of gases equates to heat, was only translational, 
thermodynamics would be much simpler since the 
interchange between heat and work would be complete 
and direct in all adiabatic processes. 

Carnot "founded" the science of thermodynamics to 
provide a rational approach to the design of steam 
engines for maximum efficiency (conversion of as 
much of available heat as possible into work). He and 
such successors as Rankine and Joule developed and 
verified thermodynamic theory which is still in use, 
with reference to empirical data, and included the "gas 
laws" of Boyle and Charles, which were also based on 
empirical data, before Maxwell and Boltzmann 
proposed the kinetic theory of gases and, of course, 
before that theory had been elaborated to include 
consideration of vibration and rotation of molecules. 

"STATES" OF MATTER (THE VAN DER WAALS' 
EQUATION OF STATE) 

Thermodynamics, as noted above, is concerned, from 
a practical perspective, with transitions of "heat" to 
"work" and vice versa. In its rudimentary form, such 
consideration involves the "gas laws", which, though 
originally based on empirical data, can be derived 
from the "kinetic theory of gases". However, as each 
of us has "always known", the gaseous state is only 
one of several in which matter exists. The first 
application of thermodynamics was to steam, which is 
the gaseous phase of water, which, when cold enough, 
freezes into ice. 

The "gas laws" of Boyle and Charles, for purposes of 
"engineering thermodynamics" are usually combined 
in the "ideal gas equation": 

PV = nRT (3) 

where: 

P is pressure, 

V is specific volume, 

n is the quantity (gram moles) of gas, 

R is the "gas constant", and 

T is the absolute temperature. 

Although "live" or "dry" steam behaves as an "ideal 
gas", at lower temperatures and higher pressures, 



steam condenses to liquid water and the "gas laws" 
cease to apply. 

Van der Waals accounted for this changed behavior 
with the change from the gaseous to the liquid state in 
terms of the Maxwell-Boltzmann "kinetic theory of 
gases" by assuming an attraction (a) between, and a 
finite volume (b) of molecules, to derive (in 1873 (5)) 
his "equation of state": 



(P+a/V 3 )(V-b) = nRT 
as given in Ref. (11). 



(4) 



Considered from the intermolecular perspective, the 
van der Waals equation of state implies that, as two 
molecules approach one another they are mutually 
attracted by a force which varies inversely with their 
separation until they collide and bounce apart "like 
tiny billiard balls". After bouncing apart, each 
molecule flies until it bounces off another. If the 
temperature T, is above the "boiling point", the 
molecules behave in accordance with Maxwell- 
Boltzmann "kinetic theory of gases" and, of course, as 
an "ideal gas" in accordance with the "gas laws". At 
lower temperatures, their kinetic energy is insufficient 
for each molecule to escape the attraction of its 
neighbors before it is bounced back by collision with 
a molecule of the gas. This effect, at liquid gas 
interfaces, is known as "surface tension". 

As it seems from casual observation and all but the 
most precise empirical data, water (as well as other 
liquids) is considered in hydraulics and hydrodynamics 
to be of constant density. Although more precise data 
have shown water and other liquids to have finite 
compressibilities and thermal expansion coefficients, 
the fact that they are considered to be negligible in 
most practical applications is evidence that the 
amplitude of the molecular motion, apparent as "heat", 
is small compared with the gross linear dimensions of 
the molecules, so that it can be considered "vibrational 
heat" like that of the relative movement of atoms, 
ions, and other groups of atoms within molecules as 
discussed above. 

Considered from a closer perspective, the analogy of 
molecules to "billiard balls" is seen to be a simplifying 
approximation which applies to gases and liquids at 
temperatures high enough that the amplitudes of the 
vibrations of "vibrational heat" are large compared 
with the deviations from spherical symmetry of the 
molecules. At lower temperatures, the weaker 
vibrations (which are still random) result in repeated 



348 



"trials" (reorientations) until adjacent molecules "fit 
together" and move closer, with a resulting increase in 
their mutual attraction (the "van der Waals force"), 
which, along with their asymmetry, maintains their 
relative positions and orientations. This is the process 
we observe as "freezing", "solidification", or 
"crystallization". 

Each of us has been aware since early childhood that 
water boils and freezes, and as our vocabularies 
increased we learned that "water", "ice", and "steam" 
are three "phases" or "states" of the same substance. 
With the passage of time, we found out that most other 
substances also can exist in "gaseous", "liquid" and 
"solid" states. 

Our earliest memories (not quite earliest for those in 
my age group) include the sight of "neon signs" 
glowing on store fronts, billboards, etc. Later, we 
heard or read that the neon lights were tubes filled 
with rarified gases, (neon, in the originals, which glow 
red). Other gases glow in other colors, but they are 
all called "neon signs" which glowed when "ionized" 
by the flow through them of electric current. In this 
context, we learned, as we became more sophisticated, 
that "ionization" meant the separation of electrons 
from atoms (or the ions of molecules from one 
another), after which, in an electrical field, the 
electrons (and negative ions) are attracted to the 
"anode" (positive electrode), and the positive ions 
(atoms or groups of atoms with electrons missing) are 
propelled toward the cathode (negative electrode). If 
the electrical field is sufficiently strong, and the free 
paths are long enough, the ionization is maintained by 
collisions, which "knock" more electrons free as 
others enter atomic orbits with resulting luminosity. 
Gases are also ionized by temperatures high enough 
that the "vibrational heat" is sufficient to overcome the 
attraction between their ions and collisions between 
molecules "knock them apart". Ionized gas is referred 
to as "plasma", and as a "fourth state of matter"(9). 

Plasma glows. Its light emission be can seen, from 
the "intra-atomic" perspective, to result from the 
falling of electrons into orbit. The color, wave length, 
or spectral characteristics of the light are unique for 
each element, and are used in spectroscopic analysis, 
to identify them The familiar blue flame of a gas stove 
results from such luminosity of plasma of which 
carbon and oxygen are principal constituents, and the 
red light of a railroad "fiizee" is the spectral emission 
of strontium, 



Van der Waals (in 1873) considered the relationship 
between temperature, pressure, and specific volume of 
substances close to their "boiling points" from the 
perspective of the recently (1871) presented 
Maxwell-Boltzmann "kinetic theory of gases" , 
assuming a finite size of each molecule and an 
attraction between similar atoms and molecules which 
became known as the "van der Waals force". A half 
century or more later, from the intra-atomic 
perspective of the "Bohr atom", the "van der Waals 
force" was seen to be "caused by a temporary change 
in dipole moment arising from a brief shift of orbital 
electrons from one side to another of adjacent atoms or 
molecules" (16). 

The van der Waals equation of state was derived as a 
basis for thermodynamic analysis of systems involving 
"wet steam" in which water is present in both gaseous 
and liquid states. The "van der Waals force" is also, 
as discussed a few paragraphs back a factor in 
crystallization, the transition from the liquid to the 
solid state. However although it plays a role, the "van 
der Waals force" is not all that holds crystals together. 
Other forces, visible from the "interatomic" and 
similarly close perspectives, contribute. Thus though, 
as would be expected (based on the description of 
crystallization a page back, in which the "van der 
Waals force" was invoked), most substances "shrink" 
when they solidify, water does not. My impressions, 
from "intra-atomic", "inter-atomic" and 
"intermolecular" perspectives are too "fuzzy" to 
include as a logically cohesive part of this paper, but 
a few seem sufficiently relevant to mention. Some 
molecules, particularly those of a number of "organic" 
compounds, are so large and/or complex that they 
exist only in the solid state. They "decompose" (break 
into smaller molecules) at rates dependent on 
temperature and the "reaction kinetic" properties of the 
substance. Conversely some compounds which are 
liquid at "room temperature", "polymerize" (their 
molecules join to form larger ones, which exist only in 
the solid state) when heated or "catalyzed" and they 
solidify. 

Crystalline and chemical bonds are similar in that they 
are effects of and governed by forces and principles 
discernible from "intra-atomic", "inter-atomic", and 
"inter-molecular" view-points and that, from the 
empirical perspective, they are "exothermal" (evolve 
heat) (because solidification prevents translational and 
rotational movement of atoms and molecules so that all 
thermal energy becomes "vibrational heat", which is 
"sensible"). 



349 



Purity is a relative term. The word "pure" is often (in 
some contexts, usually) preceded by a percentage. 
"100% pure" is not quite credible. So, all substances 
are mixtures. Each solid and liquid has a finite vapor 
pressure (or gaseous decomposition product) and thus 
an odor, which may not be apparent to most humans, 
but is to many animals and can be detected by means 
of spectroscopy. Similarly, gases and solids are soluble 
in liquids and gases and liquids are absorbed or 
adsorbed by solids. The distinction between the states 
of matter is, thus, based on practical and empirical 
considerations. 

Many familiar substances, including wood, whipped 
cream, mud, smoke, mashed potatoes, glue, wet- 
cement, and shaving cream are composed of matter in 
more than one state and owe their characteristic 
properties to interactions of their components in two or 
more states. The properties of such mixtures depend, 
to some extent, upon the "state of aggregation" (the 
size, shape (often fibrous), hardness, frictional 
properties, and distribution of solid components, and 
the sizes and distribution of droplets, bubbles, and 
pores, and, where such mixtures seem solid, from the 
"empirical" perspective, upon the structure of the 
substance, including the bonds (molecular, crystalline, 
and other) between the components, as well as their 
distribution. 

Matter, in its various states, has been viewed from 
"practical" and "empirical" perspectives. With a view 
to prediction of the behavior of systems, empirical data 
are plotted, and the indicated relationships are 
expressed in algebraic equations, which, when used in 
analysis with calculus or differential equations, are 
assumed to apply to infinitesimal intervals, an 
assumption whose validity, is questionable when 
considered from the "intermolecular" or other 
theoretical perspectives which have been discussed 
herein, but have been essential to the advance in the 
states of the arts to which they apply, and, where the 
principles of logic and mathematics have been applied 
to the satisfaction of the scientific community, have 
come to be accepted as "rigorous theory". 

Although the van der Waals equation of state (equation 
(4)) relates pressure (P), specific volume (V), and 
absolute temperature (T) for substances under 
conditions where both gaseous and liquid states exist, 
the phrase "equation of state" seems to have acquired 
the more general meaning of any equation relating 
specific volume to pressure, such as that for adiabatic 
compression or expansion of an gas"; 

PV 7 = a constant (5) 



which has been referred to as the "gamma (y) law 
equation of state" , which is considered to be one of the 
"ideal gas laws" derived from the empirically 
established laws of Boyle and Charles, which can be 
derived from the Maxwell-Boltzmann "kinetic theory 
of gases"(17) The "wet steam", to which the van der 
Waals equation of state was first applied, is a 
"colloidal suspension" of liquid water in gaseous 
steam. As has been mentioned, "colloidal 

suspensions" consist of droplets or particles of liquids 
or solids which are too tiny to settle out from the fluid 
in which they are suspended. Qualitatively, their 
failure to settle out can be ascribed to their 
bombardment from all directions by molecules close to 
their size and (in the case of "wet steam") of a similar 
density. A quantitative explanation is beyond the scope 
of this paper. However, the observation, in 1827, by 
Brown, of what came to be known as "Brownian 
motion" of colloidally suspended cells and particles, 
was explained on these bases (in 1871) by the 
Maxwell-Boltzmann "kinetic theory of gases" and 
elucidated by the concept of "Maxwell 1 s demons" (5). 

It is my impression that the phrase "equation of state" 
was first used to identify the relations between 
pressure (P), temperature (T), and specific volume (V) 
of "wet steam", a colloidal suspension of liquid water 
in gaseous steam. In the "gamma equation of state" in 
its original application, to "ideal gases", the effect of 
temperature is taken into account by the use of 
"gamma" the ratio of specific heat at constant pressure 
to that at constant volume (= C p /C v ). 

For hydrodynamic consideration of the behavior of 
substances in other states or "states of aggregation", 
experimentally determined relationships of specific 
volume (V) to pressure (P), referred to as "equations 
of state" (often "gamma law" equations of state" with 
empirically determined values of gamma) are used. 

From the practical perspective of the designer of 
hydraulic systems, water and other liquids are seen as 
incompressible fluids. However, in consideration of 
large or sudden changes of pressure (particularly, 
those of detonation and the strong shock waves it 
induces), their compressibility must be taken into 
account (11). 

To some, the mention of detonation in the above 
paragraph may seem to be a change of subject from 
that of "fire" indicated in the title hereof. It's not, 
because detonation is a form of combustion as will be 
pointed out in the section hereof headed - 
EXPLOSION AND DETONATION. However, it may 



350 



be somewhat premature at this point, so farther 
discussion is postponed until we get to it there. 

The subject of detonation came up at this point because 
consideration of detonation from a theoretical 
perspective involves relationships between pressure 
and specific volume which, as mentioned above, have 
come to be known as "equations of state", even for 
porous solids, where materials in more than one state 
are present. For such materials, a more appropriate 
term for the pressure/volume relationship might be 
"equation of state of aggregation" (which doesn't "roll 
off the tongue like "equation of state") 

Fire, usually, involves changes of state. Yellow flames 
of candles and wood and trash fires are glowing black 
smoke (a colloidal suspension of carbon particles 
(soot), the result of evaporation and condensation of 
the fuel or volatile components or decomposition 
products thereof and the subsequent decomposition of 
this colloidally suspended condensate to carbon and 
gaseous products whose oxidation provides the heat 
which sustains the process. Such fires involve 
transitions from solid to liquid to gaseous states, 
followed by reversals to the liquid state (in a colloidal 
state of aggregation), and finally back to the gaseous 
state before the oxidation takes place.. The familiar 
blue flame light of a gas stove flame is spectral 
emission of the plasma to which the gaseous products 
of combustion have changed. Glowing coals glow due 
to the oxidation of solid carbon to gaseous carbon 
dioxide. 

That changes of state occur at specific temperatures 
was common knowledge long before quantitative scales 
of temperature were proposed. Both Fahrenheit and 
Celsius established their "degrees" as fractions (1/1 80th 
by Fahrenheit, and l/100th by Celsius) of the difference 
between the freezing and boiling points of water. The 
scales having been established, and instruments for 
measuring temperature having become available, 
determinations were made of freezing and melting 
points of other substances. Since it was known that 
fuels started to burn when heated sufficiently, it 
seemed reasonable to determine the "ignition point" or 
"kindling point" (the lowest temperature at which a 
substance will continue to burn without addition of 
external heat (16)) of each of various substances. For 
most fuels, which require air, oxygen, or some other 
oxidant for combustion, such determinations present 
experimental difficulties. Pressure had been found to 
affect freezing and boiling points and, it was 
suspected, might affect kindling points. Where the fuel 
was solid and the oxidant gaseous, the temperature and 



fiber stress of the fuel and the temperature and 
pressure of the oxidant could interact to affect ignition. 
These experimental difficulties seem to have resulted 
in such skepticism on the part of their editors 
regarding "ignition points" that none of the standard 
handbooks (10,11,17) at hand as I write this, includes 
such data. These difficulties are alleviated for 
substances or mixtures which contain or include 
oxygen or another oxidant or which react 
"exothermally", (with the evolution of heat). Thus the 
Smithsonian Physical Tables (18) include tables of 
ignition temperatures of gaseous and dust mixtures 
with atmospheric air of several fuels and several 
publications (4, 19, 20, 21) include "ignition points" 
or "explosion temperatures" of pyrotechnics and 
explosives. 

Although my doubts regarding the concept of an 
"ignition point" as a physical property of each fuel 
dated from my childhood observation of "spontaneous 
combustion", the general concept by those with whom 
I discussed such matters, (and apparently others (6V£)) 
combined with my practical experience with "lighting" 
fires persuaded me of its general validity. I visualized, 
the propagation of fire as the progressive heating of 
the fuel, by the heat of combustion, to its "ignition 
point" . 

"Fire", the subject of this paper, has been defined (2) 
as "The visible heat and light emanating from any 
body during the process of its combustion or burning. " 
As has been pointed out, or at least implied, in the 
foregoing, heat and light are forms of energy, of 
which my changing perspectives are discussed in the 
following. 

CHANGING PERSPECTIVES OF ENERGY 

As mentioned before, herein, each of us sees things 
from a constantly changing series of perspectives. The 
following account of the succession of my perspectives 
of energy is included in the belief that it roughly 
parallels that of most who may read this, as well as 
those who have considered such matters in the past and 
whose views have been alluded to. 

My earliest impression of energy was that of a busy 
person who could stay busy all day. This perspective, 
which seems to be that of the fitness program 
participant who reported "having more energy" as a 
result of a low calory diet and an exercise program 
(which seems contradictory from perspectives I have 
gained more recently.) 



351 



As seen from this earliest perspective, play required 
energy. The experiences which came with play, 
coasting down hill, bouncing balls, etc., lent reality to 
views of energy from perspectives to be gained in 
school and elsewhere. 

Work, like lawn mowing, also took energy, and after 
such work, I had less energy left for other activities. 
After school work, on the other hand, I had more 
energy for play. 

In ninth grade "General Science", I began to acquire 
"physical "perspectives of energy. "Work" was 
defined as a form of energy equal to force times 
distance, which was transformed to other forms of 
energy as it was accomplished. For example, the 
work of lifting a pound weight a foot was transformed 
to one "foot-pound" of "potential energy". If the 
weight is dropped, the potential energy is transformed 
to "kinetic energy". 

My science teach, aware that we were also taking 
algebra, taught us equations for work (W): 

W = Fx 

where F is force and x is distance, 

potential energy (PE): 

PE = mhg = wh (6) 

where m is mass, w = mg is weight, h is 
height, and g is the acceleration of gravity, 



and kinetic energy (KE): 
KE = mv 2 
where v is velocity. 



(7) 



Since we were familiar with the English system of 
units, he told us that work (w) was measured in foot- 
pounds, force (F) in pounds, distance (x) in feet, mass 
(m) in "slugs" (a mass of one slug weighs 32 pounds) 
and velocity (v) in feet per second. Thus, I began to 
see things, including energy, from "analytical" and 
"mechanical" perspectives. 

He told us that, as a weight falls, the sum of its 
potential energy and kinetic energy remains constant in 
accordance with the "law of the conservation of 
energy". He demonstrated the conservation of energy 
with a pendulum, which continued to swing, 
alternately transforming potential energy to kinetic 
energy and kinetic energy to potential energy. He 
explained the reduction of the swing as the result of 



friction, which, he said, transformed the kinetic energy 
to heat, which, he said, is another form of energy. 

As its name implies, "General Science" includes many 
subjects, including those mentioned above and heat, 
chemistry, electricity, waves, (gravity (on water 
surfaces), elastic (including sound), and 
electromagnetic - radio, light, etc.), radiation (usually 
electromagnetic waves, but sometimes streams of such 
particles as electrons, protons, etc.). Each of these 
subjects deals with one or another form of energy 
and/or transformations of one form of energy to 
another. The conservation of energy was shown (from 
perspectives assumed to be familiar to ninth graders) 
to apply to all transformations from one form to 
another. 

Play, experience, conversation, observation, and 
recreational reading extended the range of my 
perspectives, some of which have been discussed 
herein before. Some early impressions, like that (from 
an Aristotelian perspective) that heavy objects fall 
faster than light ones, were corrected in the above 
mentioned "General Science" course. It was explained 
that air friction slows light objects more than it does 
heavy ones. Similarly a sled, coaster wagon, or 
bicycle is slowed less by friction on a steep hill than 
on a gentle slope, so it goes faster. If it weren't for 
friction, the speed, after a given change in elevation 
would be unaffected by the slope, since all of the 
potential energy would have been transformed to 
kinetic energy. 

As I advanced through high school, geometry (both 
plane and solid), trigonometry, and advanced algebra 
provided "graphical" and "analytical" perspectives, 
and chemistry and physics provided "scientific" 
perspectives, from which I could reexamine 
impressions gained, since early childhood, from such 
previously mentioned perspectives as "eyewitness", 
"common sense", "intuitive", and "practical". 

The combination and interaction of experience, 
observation, and education persuaded me that work, 
heat, light, and sound are forms of energy and such 
forces as those of gravity, and magnetic and 
electrostatic "attractions of opposites" are factors of 
energy, as are time and distance, and that, in any 
isolated system, the total energy remains constant 
(which is "the law of the conservation of energy") 

The freshman physics course, which was required for 
all M.I.T. students, in which the notation of calculus 
(also required) was used to express Newton's laws of 



352 



motion and gravity, and to derive from them equations 
of motion of falling bodies, basic principles of 
ballistics, and the equations of orbital motion of the 
planets, as Newton had nearly 300 years before, 
showing that Kepler's Laws, which were 
generalizations of Brahe's observations, were empirical 
verification of his laws of motion and gravity. 

The lecturer of the course Nathaniel H. Frank, who 
was also the author of the textbook "Introduction to 
Mechanics and Heat" (22), used in the course, which 
included consideration of the language of physics and 
unit systems (metric and English), kinematics (both 
linear and plane - introducing the concept of vectors), 
kinetics, and statics of mass points and particles 
(including Newton's laws of motion and gravity - 
planetary motion is considered from Copernicus' and 
Kepler's extra orbital perspective, from which the 
planets are seen as mass points), linear and plane 
dynamics, work and energy, potential energy, 
hydrostatics, elasticity, acoustics, heat conduction, 
thermodynamics, the first law of which is the 
conservation of energy, which it shows to be 
applicable in gases to adiabatic systems (from which 
no heat is lost), as well as to reversible mechanical 
processes (as distinguished from irreversible processes 
such as frictional heating).The text discusses "entropy" 
(S), a term coined by Clausius (5) for the ratio 
(S = Q/T) of the heat content (Q) of a system to its 
absolute temperature (T), which was shown to be a 
measure of the unavailability of the heat for 
transformation to work, and quotes Clausius statement 
of the first and second laws of thermodynamics in 
closing, - "The energy of the universe remains 
constant. The entropy is always increasing. " 
Recently, in retrospect, I have wondered how I 
reconciled this statement with my impression, from the 
"cosmic" perspective gained in recreational reading, 
that the sun and other stars had been radiating energy 
for billions of years. Perhaps Frank had cited 
Einstein's "Special Theory of Relativity", which holds 
that mass (M) is a form of energy (E) which is 
expressed in: 



E = Mc 2 

where c is the speed of light. 



(8) 



I do remember having read, a few years earlier, that 
the energy radiated by stars was the product of 
reactions of atomic nuclei and that there was enough 
energy in a glass of water to propel an ocean liner 
across the Atlantic, which had led me to envision a 
device capable of transforming nuclear energy into 
work, which would fit into the rear hub of a bicycle 



(like a coaster brake). I had no idea as to how this 
might be accomplished, but thought it would be nice 
to have one installed in my bike. My freshman year 
was 1934-'35. A decade later, the conversion of 
nuclear to other forms of energy (on a much larger 
scale) was an important factor in the conclusion of 
World War II. I still have no idea as to how it might 
be applied to bicycle (or even automobile) propulsion, 
but it is now used to propel submarines and generate 
electric power, some of which has been used to charge 
the batteries of electrically propelled cars. However, 
in the 1930s, nuclear energy was considered to be the 
"stuff of science fiction" (like space travel) and 
engineering courses were concerned with more 
"practical" matters. 

Another freshman course was Synthetic Inorganic 
Chemistry. 

Sophomore courses included physics (optics, 
electrostatics, electrodynamics, and magnetism) 
differential equations, machine drawing, physical 
chemistry , graphic analysis , and industrial 
stoichiometry. 

As a mechanical engineering major, I took courses in 
applied mechanics (including stress analysis, 
kinematics, and kinetics), metallurgy, fluid 

mechanics, materials testing, manufacturing and 
construction processes, as well as such "general" 
subjects as English, history, descriptive astronomy, 
and economics. 

Each course considered its subject from a unique 
perspective, each of which was somewhat familiar to 
me from earlier education, experience, and reading, 
and some of which have been mentioned herein. From 
one perspective it is apparent that each form of energy 
is either potential or kinetic energy or a combination 
thereof, and that other forms of energy, such as heat, 
sound, electromagnetic radiation (including light), etc. 
are manifestations of these. Each classification is the 
result of the perspectives from which it has been 
considered. From the "inter molecular" perspective of 
the Maxwell-Boltzmann kinetic theory of gasses, for 
example, heat is seen as kinetic energy of molecules, 
ions, and atoms. From the "interatomic" perspective, 
which has been discussed, it becomes apparent that, 
while this view is applicable to "translational" heat, 
"rotational" and "vibrational" heat are combinations of 
kinetic and potential energy as are sound and vibration 
as well as gravity waves on water surfaces. An 
electrical charge is potential energy while "direct 
current" is kinetic energy of electrons and "alternating 



353 



current" alternates between kinetic and potential 
energy. The "heat of combustion" of fuels, in general, 
is the potential energy of the attraction of the atoms 
and ions of the carbon, hydrogen, and other elements 
with positive valences, which they may contain, to 
those of oxygen. 

Consideration from the several perspectives discussed 
herein has left me with the conviction that physical and 
chemical phenomena and processes are, in general, 
transformations of energy between forms, often 
involving changes of state of the matter involved. 
Although views from several perspectives, which have 
led me to this conviction, have been discussed 
previously herein, the following account of my 
progress toward it (as recalled decades later) may tend 
to substantiate the conviction in the minds of readers: 

My earliest quantitative impressions related to energy 
were in terms of power (which, I was to learn, means, 
in general, the rate at which energy is transformed 
from one form to another). Light bulbs were (and are) 
graded in watts, a unit of power. Then, as now, 
illumination of a room or other space was quantified, 
by many, in terms of "watts of light". I'm still not 
sure that those who refer to light in these terms are 
aware that the watt is a unit of power and I doubt that 
many recognize (as I have come to with the passage of 
years) that the rating of a light bulb in watts is a 
statement of the rate at which it is expected to 
transform electrical energy, by "ohmic heating" into 
heat, which is, in turn transformed, as "black body 
radiation" into electromagnetic radiation, mainly in the 
visible range of the spectrum. My earliest impressions 
of the relative power of automobiles and outboard 
motors came from advertisements of their 
"horsepower". I heard (or read) that James Watt had 
coined the term "horsepower" for use in 
advertisements of the steam engine he had invented in 
terms which, he hoped, would appeal to his intended 
customers. In 1783, based on experiments with a 
strong horse, he established the value of a horsepower 
as 550 foot pounds per second. By 1800, the metric 
system, which included the watt, so named in honor of 
Watt, who had defined "power" as a physical quantity 
(a horsepower is 746 watts), had been accepted by an 
international commission and has since been adopted 
internationally by scientists. Although most ratings of 
devices which are activated by electricity seemed to be 
in terms of power, it was paid for by the "kilowatt 
hour", a unit of energy. Of course, a kilowatt hour is 
more than 2 x /i million foot-pounds (the foot pound was 
the first quantitative unit of energy I learned about in 
school). 



A difficulty in the consideration of transformations of 
energy in quantitative terms is the variety of units in 
which physical quantities (including energy) are 
expressed. The above discussion of transformation of 
energy between mechanical forms is a repetition of the 
explanations, (as I remember after sixty-some years) 
by my science teacher, who used the English system, 
with which we were familiar. It seemed reasonable 
that it took a foot-pound of work to lift a pound a foot. 

I learned, in ninth grade "General Science", that the 
work of lifting an object weighing a pound a foot was 
transformed into a foot-pound of potential energy and 
that, if the object was dropped, the potential energy 
would be transformed into kinetic energy. All of which 
seemed to confirm my previous experience and the 
aphorism that: "What goes up, must came down.", 
which has been applied, with varying degrees of 
pertinence, to prices, temperatures, unemployment, 
voltages, and the popularity of entertainers. I had also 
noticed that some things, when lifted and dropped on 
to same surfaces, bounced, and/or made a noise when 
they hit. With the passage of time, I learned to explain 
the bounce as the result of the transformation of 
kinetic energy to elastic potential energy followed by 
its transformation back to kinetic energy, and the noise 
as the result of the transformation of same of the 
energy to sound (which is alternately kinetic and 
potential energy). Nothing seems to bounce forever, 
because, I learned, at each bounce, same of the kinetic 
energy is transformed into sound and some is 
transformed into heat. The foregoing seemed a 
satisfactory qualitative explanation, but a 
demonstration in quantitative terms was difficult, not 
only because of the problems of measuring the 
quantities of sound and heat evolved during a bounce, 
but also because of the problems of conversion 
between the units in which the results of such 
measurements would be expected and those in which 
kinetic energy is expressed. (Energy has been 
expressed, by specialists in various fields, in 
foot-pounds, inch-ounces, foot-tons, BTUs, ergs, 
joules (watt-seconds), watt-hours, kilowatt-hours, 
calories (cal.,(gm)), and Calories (cal.,(kg) or 
kilocalories). Some scientists express the view that 
confusion can be eliminated by the use of the 
"universal" metric system, Perhaps, but a Ph. D. 
chemist once told me of an instance when he and a 
colleague allowed the ice cubes in their drinks to melt 
in their mouths, assuming that this would absorb the 
calories in the alcohol , forgetting that some 
nutritionists refer to Calories as "calories" (so they'd 
have had to melt two kilograms, rather than two grams 
of ice to absorb the 150 nutritionists "calories" in each 



354 



drink). In view of these difficulties, I satisfied myself 
with consideration of this matter in terms of the 
"coefficient of restitution" the ratio (e=v 2 N x ) of the 
(upward) velocity (v 2 ) of an object after it bounced to 
its (downward) velocity (Vj ) before it hit. 

The transformations of energy, mainly mechanical 
forms (work, kinetic, and potential energy are 
considered above, from "eyewitness" , "common 
sense", and "empirical" perspectives). As a student of 
mechanical engineering, I acquired a thermodynamic 
perspective, from which transformations heat and work 
are considered and that of applied mechanics (which 
considers relationships of stress and strain, the integral 
of which is elastic potential energy). Other courses, 
which are mentioned a page or two back, presented 
hydraulic, hydrodynamic, aerodynamic, graphical, 
analytical, kinematic, dynamic,, and stoichiometric, 
and other chemical (including organic) perspectives. 

Recreational reading had, from early childhood, 
provided a succession of perspectives, including those 
of nursery rhymes, Bible stories, Aesop's fables, fairy 
tales, Indian legends, Greek and Norse mythology, 
history, geology (24), astronomy (22), cosmology 
(23) , and atomic and intra-atomic physics (9) . 
Considered from those perspectives which seemed 
"scientific" to me, in about 1940, I saw (and still see, 
with a few revisions based on what I've learned since 
then) energy transformations in the universe and the 
world about as follows: 

Technical discussions should follow logical or 
chronological sequences, preferably both. The 
sequence of my perspectives, as remembered after fifty 
years, is neither. If I have failed in the following 
effort to put them in "proper" order, I hope that 
readers will be tolerant. 

In 1940, I was unaware of the "big bang" theory of 
the origin of the universe, which had been proposed 
(by Le Maitre (5)) in 1927. (I was to learn, from 
Gamow, of the theory, a few years later.). However, 
I had been aware and accepting of the 
Chamberlin-Moulton theory that the planets, (including 
the earth) of the solar system were the "drops" formed 
in the "breaking" of a tidal wave raised from the 
surface of the sun to a connecting arm by a passing 
star, each of which was drawn together by gravity. In 
the contraction, gravitational potential energy was 
transformed into kinetic energy, which was, in turn, 
transformed into work, and then, to heat, some of 
which was radiated as black body radiation, cooling 
the planets until solid surfaces were formed, the 



surface temperature of each planet continued to drop 
until equilibrium was reached between the radiant 
energy received from the sun and that lost by radiation 
from the planet. As each planet acquired an 
atmosphere by diffusion and volcanic eruption from its 
interior and by gravitational attraction of interplanetary 
gases and "solar wind", the temperature, in each case, 
was affected by absorption of radiation by atmospheric 
gases (referred to, in recent years, by 
"environmentalists", as the "greenhouse effect"), by 
the point to point variation of the "albedo" 
(reflectivity) of planetary surfaces, in combination with 
the rotational and orbital movement of each planet 
about non-parallel axes, has resulted in variations of 
surface and atmospheric temperatures with time and 
location, which result in the phenomena referred to as 
"weather" and "climate". The water vapor which has 
been a significant fraction of the earth's atmosphere 
has contributed to the complexity of these phenomena 
since the range of temperatures with include the mean 
equilibrium temperature of the earth's surface and 
atmosphere is conducive to the existence, of significant 
fractions of this water in each of all three (gaseous, 
liquid, and solid) states or phases, transitions between 
which are accompanied by mutations between 
translational, vibrational and rotational heat, which are 
seen from the "empirical" and "thermodynamic" 
perspective as "latent heats of fusion and 
vaporization", and changes of specific volume in 
which heat is transformed into work and consequent 
convection which transforms work into the kinetic 
energy of wind. 

The liquid water, which covered most of the earth's 
surface, dissolved some of the gases and solids with 
which it came into contact to become a "primordial 
soup" in which many chemical reactions were bound 
to take place, producing a wide variety of chemical 
compounds of varying complexity. It has been 
postulated that, given "enough time", a molecule 
would form which would be capable of reproduction 
and have the other characteristics of a living cell, and 
that such cells would further organize themselves and 
adapt to their environments to evolve to the many 
organisms which have existed on the earth. Statistical 
calculations (in the 1950s), which are cited by 
"creationists", indicate that there hasn't been "enough 
time" . More recently, numerical models of 
"coevolution", have reduced estimates of "enough 
time" (26). 

I have yet to acquire mathematical or computational 
techniques or perspectives from which I can consider 
with confidence the relative validity of the views of 



355 



"evolutionists" and "creationists", but, based on 
paleontolological evidence, I am persuaded that living 
organisms have existed on the earth for billions of 
years, which implies the presence of liquid water, and 
that the equilibrium between radiant energy received 
(less than 0.05% of the sun's radiation) and lost during 
this period, would require radiation by the sun of a 
quantity of energy which is credible only on the basis 
of the consideration that (as stipulated in Einstein's 
"Special Theory of Relativity") that mass (M) is a 
form of energy (E) as related by: 

E = Mc 2 (8) 

where c is the speed of light. 

My view of energy transformations, past, present, and 
anticipated, which seem relevant to the subject of this 
paper, is outlined below: 

Mass is transformed, in the interior of the sun and 
other stars, by thermonuclear reaction, to heat, which 
is transformed, at or near the surfaces of the sun and 
stars to "black body (electromagnetic) radiation". A 
fraction (referred to as the "albedo" of the planet) of 
the electromagnetic radiation which is intercepted by 
each planet is reflected. Most of the rest is 
transformed into heat, and, eventually, reradiated as 
"black body radiation" (maintaining its surface 
temperature equilibrium) . Some of the radiation 
intercepted by the earth, is transformed, by 
photochemical reactions (including photosynthesis) into 
(chemical) potential energy, which, for the organic 
compounds synthesized in photosynthesis which are 
used as fuels, is referred to as their "heat of 
combustion", and when they are used as foods as their 
(nutritionist's) "calory content". In animals (including 
humans) the potential energy in foods referred to as 
"nutritionist's calories" is transformed into work and 
heat by movement and metabolic processes. Fire 
transforms the "heat of combustion" of a fuel into the 
kinetic energy of molecules referred to by some as 
"sensible heat". 

Although neither prehistoric man nor I (as a kid) 
considered such matters from these perspectives. 
Transformations of energy, by fire from chemical 
potential energy ("heat of combustion") to heat, and 
from heat to light, sound, work, and kinetic energy 
have been applied to form the basis of most "know 
how", trades, technologies, crafts, arts, and sciences, 
and provided a series of perspectives, to and by 
mankind through prehistory and history, and to (by) 
me in the course of growing up, education, 
recreational reading, and research, both literature and 



experimental, of the world and everything in it, and of 
all that has happened. 

The word "efficient" seems to have originally, meant 
"effective". With the development of systems for the 
transformation of energy from one form to another, 
when preceded by a percentage, it has come to mean 
the percentage of available energy which has been 
transformed as intended. 

MY PRE-'41 PERSPECTIVES OF FIRE 

As mentioned earlier in the section headed: "A KID'S 
PERSPECTIVE OF FIRE", my earliest impression of 
fire was that of a yellow flame, which I generalized to 
the view that fire and light are aspects of the same 
thing. Grown ups talked about "lighting" fires and 
about "firelight", usages which I adopted. 

Anyone who has tried will recognize the difficulties of 
recalling how the world looked and what each word 
meant in early childhood, without having one's 
memories distorted by more recent learning and 
experience. In view of these difficulties, I'm sure that 
this account in not completely accurate (for example, 
in the final paragraph of the previous section headed 
"LANGUAGES AND PERSPECTIVES" I quoted, 
adults, explaining that the visible "steam" from a 
teakettle was not steam, but a suspension of droplets of 
liquid water, I included the words "colloidal" and 
"aerosol", neither of which are defined in terms 
applicable to the explanation in a 1939 dictionary (2), 
so they couldn't have been included in an explanation 
to me in the 1920s), but it is the best that I can do. 

I could see the light and feel the heat of a fire and of 
the sun, and got the idea that light and heat are 
related, but not the same. If the fire was in a stove, I 
couldn't see its light, but could feel its heat and, 
though I could see sunlight reflected from snowbank, 
I didn't feel much heat. 

In time, I began to see that heat is needed to start a 
fire and that fire is a source of heat. The heating 
elements of other sources of heat, like electric stoves, 
toasters, and space heaters, glowed when they were 
hot enough, and I began to recognize that the light of 
a flame was an effect of its heat. After seventy some 
years, I can't remember my introduction to the concept 
of an "ignition" or "kindling point" as a property of 
each fuel, but I do remember my observation (which 
is described in that earlier section) of "spontaneous 
combustion", which was the source of my reservations 
regarding that concept. Practical experience induced 



356 



me to accept the concept, with the reservations alluded 
to. Paper was easy to light with a match, apparently 
because its thickness was small compared with the 
dimensions of a match flame so that some of it could 
heated to its 'kindling point". As a Boy Scout, learning 
to build a fire without paper, as required to pass the 
second class firemaking test, I was taught to use twigs 
or shavings of dimensions similar to those of a match 
stick to pick up the flame from the match. It took a 
few seconds, apparently, to heat the twigs to their 
kindling point. Once the twigs were burning, larger 
pieces of wood were placed in the flame, The larger 
sticks took longer to "catch fire", as I saw it, because 
there was more wood to heat to its "kindling point". It 
became apparent that the ignition and spread of fire 
depends upon the heating of the fuel to its kindling 
point by an external source of heat or the established 
fire. 

I was told, when building a fire, to place the new 
sticks or logs, above those which were already 
burning, because "Heat rises.". As I grew older, I 
learned that the effective rise of "heat" was, more 
accurately, stated as "Hot air rises. " due to 
convection, which occurs because fluids, including air, 
expand when heated, and become buoyant with respect 
to fluids of similar composition (but cooler.) I learned 
that other heat transfer mechanisms were conduction 
and radiation. 

Like, I suppose, most people, both living and dead 
(some for long times), I had experienced heat transfer 
by all three mechanisms since early childhood. My 
early experiences with light and "radiant heat" (which, 
I learned , after a few years , is called " infrared 
radiation" in the language of physics), are recalled a 
page back. We have all felt the heat conducted from 
warm and hot objects. In the kitchen, heat is 
conducted by a frying pan, from the burner to the 
food, in toasting and broiling, heat is transferred by 
radiation. Boiling and baking involve convection. 

Convection is utilized in a hot air heating system, to 
transfer the heat, from the furnace in the basement, 
through a duct system to living quarters on the floors 
above, and, for fireplaces, and coal and wood furnaces 
and stoves to provide the "draft" of air needed to keep 
the fire burning. 

My memories of youthful impressions of heat transfer 
and, more specifically, convection are outlined above. 
In the course of the recall, it occurred to me to check 
recent references regarding the current meanings of the 
words. Dictionaries, both 1939 (2) and 1976 (16) 



define "convection" essentially as described above, 
The Random House Dictionary (1) defines it as "The 
transfer of heat by the circulation of the heated parts 
of a liquid or gas", with no mention of cause of the 
circulation. The term seems to be sometimes used in 
this latter sense. 

I had been aware, since early childhood, that fire 
required air. I'd watched while people "smothered" 
fire, or blown or fanned fires to make them burn 
faster or hotter, and had been shown how to regulate 
a fire by adjusting a "damper". All of which prepared 
me for the "chemical" perspective which is discussed 
in the earlier section under that heading, and the 
recognition that the fire with which I was familiar was 
oxidation. 

As my perspectives changed between the several which 
have been mentioned in the foregoing, my views of 
fire and the changes of state and composition of 
matter, and transformations and transfers of energy 
between forms and locations involved, changed as if I 
was "channel surfing" on a television set with a 
"zapper" (to use a metaphor which would have been 
meaningless in the 1930s). On rainy days, I'd seen 
water flow down hill, faster down steeper slopes. 
When I acquired a graphical perspective, and saw it 
applied to hydrodynamics, electricity, and heat 
transfer, pressure, voltage, and temperature seemed, 
almost always, to be plotted as vertical displacements 
from the origins of graphical representations of the 
spatial distribution of these quantities. By analogy to 
water "seeking its level" I saw fluids, electricity, and 
heat flowing downward, from points or regions of high 
pressure, voltage or temperature at rates proportional 
to (and in the direction of) the gradients (or "slopes") 
of these quantities. I have recently learned that this 
view of heat transfer (as seen from the "empirical" 
perspective) led Lavoisier and, later Mach, to see heat 
as a fluid (5). 

As I remember at the time of this writing, the 
perspectives I gained from play with a chemistry set 
was more accurately characterized as an "alchemic" 
than as a "chemical" perspective. Like the ancient and 
medieval alchemists, I followed recipes and observed 
reactions. The operations of the "Universal Research 
Laboratories", which are also mentioned in that section 
were, like Roger Bacon's thirteenth century 
experiments with gunpowder (4), more alchemy than 
chemistry. 

Although I may have acquired a (somewhat indistinct) 
chemical perspective from the activities mentioned 



357 



above and recreational reading, my high school 
chemistry course clarified my chemical perspective and 
presented a few glimpses from the "intermolecular" 
perspective mentioned under that heading. 

From the "intermolecular" perspective, it became 
apparent that even so simple a reaction as that depicted 
in equation (1): 



2H 2 + 2 



2H 2 



(1) 



is not the single step reaction implied by the 
stoicheometic equation (1). Each oxygen molecule 
must be dissociated to provide the single oxygen atom 
for each water molecule. Based on models of water 
molecules, such as those shown in Figure 1, in which 
the hydrogen atoms are on opposite sides of the much 
larger oxygen atoms, it seemed that the hydrogen 
molecules also must be dissociated to be oxidized. 

I don't remember how or when, but at some time 
before I graduated from high school I became 
convinced that heat is molecular motion, the nature of 
which has been discussed herein. In the solid state, 
where crystalline bonds hold the molecules in their 
relative positions, only vibration about its equilibrium 
position is possible for each molecule (or atom). With 
increasing temperature, the amplitude of the vibration 
is sufficient to move each molecule so far from its 
equilibrium position that the crystalline bonding force 
can no longer restore it, and the solid melts. In the 
liquid state, molecules are free to move relative to one 
another, but are drawn together by the "van der Waals 
force", further increase in temperature results in 
movement beyond the effective range of this force and 
the liquid was said to "boil" or "vaporize". In the 
vapor or gaseous state, molecules are separated 
sufficiently that they move independently until they 
collide, as can be seen from the perspective of the 
Maxwell-Boltzmann kinetic theory of gases. 

Considered from the perspectives outlined in the 
foregoing paragraphs, I saw heat conduction to be the 
result of essentially mechanical interactions of 
molecules and atoms. In a solid, the vibration of each 
molecule about its equilibrium position is 
communicated to its neighbors by the same crystalline 
binding forces that establish their equilibrium relative 
positions. In a liquid, the molecular motion is 
communicated mostly by the van der Waals 
intermolecular attraction force and the intermolecular 
repulsion which determines the effective volume of 
each molecule and hence the specific volume of the 
liquid. In a gas, viewed from the perspective of the 
Maxwell-Boltzmann kinetic theory of gases, the 



movement is communicated by effectively elastic 
collisions between molecules, 

Although the view of heat transfer as seen from the 
"intermolecular" perspective, as described above, was 
more consistent with the structure of matter in its 
various states, as seen from this perspective, 
quantitative consideration would involve too much 
complex computation (since it would have to take into 
account the variation with relative directions of the 
intermolecular, interionic, and interatomic attraction 
and repulsion forces) for practical purposes. 

Like most students, I learned to consider heat transfer 
from the empirical perspective (from which Lavoisier 
and Mach had seen heat as a fluid), which is a more 
practical approach to heat transfer calculations. Such 
consideration, for engineering purposes (11), yields: 



q = k A(T,-T 2 )/x 



(9) 



where: q is the rate of heat transfer through a panel 
of area A and thickness, x, and T { and T 2 are 
temperatures on either side of the panel, 
while k is the thermal conductivity of the 
substance of the panel. 

The value of k can be determined experimentally by 
measurements of q when values of other variables are 
preestablished. 

For purposes of theoretical consideration of systems in 
which heat transfer is a factor, equation (9) can be 
generalized as a partial differential equation: 



q = -k A[dTIdx] 
and in vector notation as: 
q = -k A VT 



(10) 



(11) 



I gained this perspective of heat transfer, by 
conduction, several years after I had learned to build 
fires as a Boy Scout. From this perspective, in 
combination with the concept that each fuel has a 
"kindling point** and heat capacity, I began to see why 
the techniques I had learned as a Boy Scout were 
effective and necessary. 

A log or large piece of wood can't be "lit" with a 
match because, although the temperature of the flame 
is well above the "kindling point" of the wood, its 
thermal conductivity is much lower so that the 
temperature at the surface attains an equilibrium such 
that the rate at which heat is conducted into the wood 
is equal to that at which it is conducted from the 
flame. Paper, twigs and shavings can be lit because the 



358 



heat transferred from the flame is conducted through 
the fuel only a short distance until it reaches another 
surface from which it is conducted by air, whose 
thermal conductivity is equal to or less than that of the 
flame, so the heat accumulates in the paper, twigs, or 
shavings until the "ignition" or "kindling point" is 
reached. 

While the view of ignition, from the perspective of 
thermal conduction, outlined above, explained some of 
my experiences as a Boy Scout, they left some 
observations unexplained. From the perspective of 
"states" of matter, which are discussed earlier herein, 
it is apparent that, although most of the fuels discussed 
above are solid, the flames, like the visible "steam" 
from a teakettle, the clouds in the sky and most of the 
white "smoke" which rose from a burning pile of 
damp leaves, seemed to be lighter than air. As 
mentioned in the earlier section headed 
"LANGUAGES AND PERSPECTIVES" I was told 
that the visible "steam", white "smoke" and clouds, as 
well as fog and mist, were droplets of liquid water too 
small to settle out, and referred to as "colloidal 
suspensions" as were similarly suspended droplets and 
particles of other substances. Grey, brown and black 
smoke are such suspensions of other substances as is 
evidenced by their odor, while flames are suspensions 
of carbon which is so hot it glows. The droplets and 
particles are too widely scattered to have as much 
effect on the density of the air as the heat (from the 
fire?), so the "steam", "smoke" and clouds floated 
upward. This upward movement of flames, often 
referred to, by poets, novelists, and journalists, as 
"leaping", and in technical discussion as "convection", 
plays a role in the propagation of fire, as mentioned, 
a page or so back, in the account of my recollections 
of Boy Scout firemaking. 

I don't remember when I first heard the aphorism, 
"Where there's smoke there's fire.", but I'm sure that 
it was before my twelfth birthday that I began to 
question its (absolute) truth. I'd seen smoke coming 
from overloaded extension cords, and stop after the 
appliances which had overloaded them were 
disconnected. I'd seem pictures of smoke (identified as 
such) coming from volcanoes although I hadn't been 
told of any underground source of the air which would 
have been needed to sustain a fire. I wondered what 
smoke was. The white smoke, from burning damp 
leaves was obviously, like visible "steam", fog, mist, 
and clouds, a colloidal suspension of liquid water, but 
the grey, brown, and black smoke were something 
else. I was a few years younger, when an aunt, who 
lived in an in-town apartment, reached out of her 



window and showed me a blackened finger tip. She 
identified the black stuff on her finger as "soot" 
which, she explained, had settled out from the black 
smoke, which came from chimneys of building in 
which, she said, "soft coal" was burned. It seemed to 
me that grey smoke must be a mixture of black smoke 
and white smoke, but, from its odor, I was sure that 
all smoke contained something else. 

Consideration of the observations, experiences and 
hearsay mentioned in the foregoing paragraph from my 
developing chemical perspective resulted in views of 
fire and smoke which I found satisfying. Although I 
questioned that "where there's smoke there's fire.", 
it was apparent that, where there was smoke, fire 
could be expected. If the overload which caused an 
extension cord to overheat wasn't removed, the cord 
would soon "burst into flame". By analogy to the 
formation of the aerosol referred to as visible "steam" 
beyond the spout of a teakettle, I reasoned that 
something in the insulation of the extension cord (bare 
wires, when heated by electric current, don't emit 
smoke) must have evaporated and condensed, after 
mixing with cooler air, to form the droplets of the 
colloid referred to as "smoke". Reflecting on earlier 
experiences and observations, some of which have 
been mentioned hereinbefore, I recalled that, when 
organic substances are heated, smoke often appears 
before flame. 

The word "organic", like many others, has a number 
of meanings, the most general of which is "Arising 
from an organism. "(2). Organic chemistry is 
essentially the chemistry of carbon, which owes its 
complexity to four "covalent" bonds whereby carbon 
atoms bond with one another, and those of other 
elements to form a wide variety of molecules (aver 
15000 of which are listed in the Handbook of 
Chemistry and Physics (10)). Because carbon atoms 
combine in several ways,it is possible for different 
molecules (of compounds with different properties) to 
have the same composition in terms of the numbers of 
atoms of carbon and other elements. For this reason, 
organic compounds are usually identified by structural 
formulas which are, essentially, diagrams or models of 
their structure. It is quite apparent, from 
consideration of such structural formulas, that many 
organic molecules are too large and irregular in shape 
to bounce around like "tiny billiard balls" (as Mach 
characterized the picture presented by the 
Maxwell-Boltzmann kinetic theory of gases) but are 
more likely to break into smaller molecules. In other 
words, some organic compounds tend to decompose 
rather than evaporate when heated. (I became 



359 



somewhat dimly aware of such matters at an early age 
because of my father's involvement in the development 
and construction of "cracking stills" in which large 
molecules of crude oil were "cracked" into the smaller 
molecules needed in gasoline.) 

The familiar fuels are organic materials (in the sense 
of "Arising from an organism. "(2)). As such they are 
composites of a number of substances (Mostly organic 
compounds of carbon, hydrogen, and nitrogen, in 
solid, liquid, and gaseous states. Wood, for example 
is a mixture of cellulose, which is "made up of 
long-chain molecules (fibers) in which the complex 
unit QjHioOs is repeated as many as 2000 times" (17), 
lignin (C 41 H 32 6 ) sugars, resin, acetic acid , water, air, 
and other substances. 

When a campfire had subsided to glowing coals and, 
for one reason or another, a hotter fire was desired, a 
few sticks of kindling were laid over the glowing 
coals, In few minutes (particularly if the glow was 
brightened by blowing or fanning the coals, the 
"kindling 11 began to emit smoke, which, a minute or 
two later, burst into flame. I had noticed that the 
smoke appeared before the flame. 

Consideration from perspectives hereinbefore 
discussed, particularly the "chemical" perspective and 
that form which "states" of matter are viewed, led me 
to see the sequence of observable phenomena outlined 
above as resulting from the following sequence of 
chemical and physical events: 

The "kindling" was heated by radiation and convection 
from the glowing coals. When it reached temperatures 
conducive to such processes, volatile components of 
the "kindling" began to evaporate and nonvolatile 
components decomposed to volatile compounds which 
evaporated. The vapor, mixing with cooler air 
condensed to form the droplets of the colloidal 
suspension (or "aerosol") referred to as "smoke". 
Further heating brought the smoke to its "ignition 
point" so it "burst into flame" . 

The above description satisfied me in 1940, and it still 
does except for my continuing reservation regarding 
the concept of "ignition points" and the colloquialism 
of the phrase "burst into flame" and its implication of 
an "eyewitness" rather than a "chemical" or "physical" 
perspective. 

These misgivings were alleviated by replacing the final 
sentence (which included the dubious phrases) with the 
following continuing description: 



If and when the smoke was further heated, the gaseous 
decomposition products of the wood, such as methane 
(CH 4 ), ethane (C 2 H 6 ), propane (C 3 H 8 ), carbon, 
hydrogen, etc. oxidize, raising the temperature still 
higher, decomposing and oxidizing the compounds 
which had condensed to form the droplets of the 
colloidal suspension referred to as "smoke". 

Most decomposition products of the organic substances 
commonly used as fuels are in gaseous or plasma 
states at temperatures associated with combustion. 
The most notable exception is carbon, which is solid 
at much higher temperature. The charcoal, which is 
the most familiar decomposition product of wood is 
mostly carbon. The glowing coals which remain after 
the flames have subsided are mostly carbon, which 
continues to burn while oxygen is available. Enough 
of the heat of combustion is transferred from the 
carbon dioxide, which is the reaction product, to the 
oxygen of ambient air and unburned carbon to 
maintain the reaction and the glow, which is black 
body radiation. Similarly, the organic substance of the 
droplets of the colloidal suspension referred to as 
smoke decompose to the gaseous products mentioned 
above, and the small particles of carbon, referred to as 
"soot", whose colloidal suspension in air is called 
"flame", when it glows, and "black smoke" after it 
cools enough to stop glowing. 

As mentioned a few paragraphs above,the "smoke" 
emitted by heated wood is a colloidal suspension of 
volatile components of the wood. Their composition 
varies from one species of wood to another. Many 
have proven useful. Perhaps the best application of 
wood smoke is to the preservation of food, such as 
ham, bacon, and fish. 

The flavor of smoked meat has been sufficiently 
popular to inspire the invention of the "pit barbecue" 
on which meat is smoked while it is broiled and 
roasted, although preservation is not a consideration 
because the food is eaten while it is still hot. Hickory 
and mesquite smoke seem to be most popular for these 
purpose, as well as (in condensate form) as flavoring 
for barbecue sauce, etc. 

Other volatile components of wood, which are 
undoubtably seen as wood smoke but may not 
condense to "smoke" before they are condensed to 
liquids in stills, are turpentine, which is used as a paint 
thinner and brush cleaner, and creosote, which is used 
as a wood preservative and harsh disinfectant (17). 



360 



The last couple of pages contain descriptions of fires 
with which I was most familiar before 1941, in which 
wood or paper were the fuels, as I saw them from 
several perspectives. I was also aware of combustion 
of other fuels to which some parts of these descriptions 
are not applicable. 

I was aware that although most of the other fuels I had 
seen burning were of organic origin, they tended to 
have properties more similar to the intermediates of 
wood burning than to the wood itself. Most were 
gaseous, volatile, or colloidal suspensions, like 
components of wood smoke, or mostly carbon and 
ash (like charcoal), when visibly burning. 

I had heard coal, petroleum, and natural gas referred 
to as "fossil fuels", meaning that they were the 
remains of prehistoric organisms which had 
decomposed and been buried by such geological 
processes as volcanic eruption and sedimentation. I 
had seen the beginning of the formation of coal in the 
peat bogs of the upper midwest, many acres of spongy 
moss, where a misstep could result in a foot coated, 
almost to the knee, with dark brown rotted moss, 
referred to as "peat", which, in the United States, is 
used as fertilizer and (dried) as thermal insulation, 
packing, and "potting soil" for plants, but in countries 
where other fuels are expensive, dried peat is used, in 
large quantities, as fuel. The top layers in a peat bog 
are of relatively low density, but, at greater depths the 
older peat, which has been rotting longer and 
consolidated by the combination of increasing pressure 
and the upward diffusion of water and other low 
density (compared with that of carbon [3.51]) liquids 
and gases (most of which are products of the 
continuing decomposition (rotting) of the peat). Gases 
diffuse to the air above as "marsh gas" (which 
sometimes catches fire and is referred to as "will o' 
the wisp" or "ignis fatuus"). With the passage of 
time, the growth of the moss continues, piling up more 
peat, so that at the bottom continues to consolidate 
while decomposing, and its density, carbon content, 
and hardness increase with time and depth and the peat 
is changed, progressively, to lignite, bituminous and, 
finally, anthracite coal. 

Petroleum, like coal, is a product of decomposition 
(decay) of organic matter, in this case, marine animals 
and plants, which, because of their much lower 
oxygen/hydrogen ratio than the peat moss, which is 
mainly cellulose (a carbohydrate (or hydrate of carbon, 
which can decompose to carbon and water)) tends to 
decay to hydrocarbons. As in the formation of coal, 



the process includes the "piling up" of the matter for 
long periods of time. 

In the course of the millions of years during which 
coal and petroleum have been forming, various 
geological events and processes, including volcanic 
eruptions, earthquakes, and continental drift, have 
occurred, resulting in deformation of the earth's crust 
and the formation of mountain ranges and 
displacement of continents, in the course of which 
some of the forming coal and petroleum were covered 
by layers of rock, which sealed pockets of the gaseous 
products of the decomposition of organic matter 
whereby the coal and petroleum are formed. These 
trapped gases are known as "natural gas". 

Some fossil fuels are used as recovered. Coal is 
burned in furnaces to heat buildings, and in boilers of 
locomotives, ships, and power plants to generate 
steam. Natural gas is distributed through pipes to 
residences and other buildings where it is the fuel for 
furnaces, space heaters, cook stoves, fireplaces, and 
refrigerators. 

Some coal, generally bituminous coal, is heated in 
kilns, to continue the process of decomposition to 
gaseous hydrocarbons (methane, ethane, etc.), which 
are referred to as "manufactured gas" and distributed 
through pipes in communities beyond the range of 
natural gas distribution pipelines, and carbon and ash, 
known as "coke", which is sold as household fuel, and 
used in blast furnaces in which iron ore is reduced to 
"pig iron", some of which is remelted and cast, to 
make the wide variety of cast iron items with which 
we are all familiar, but most of which is converted to 
steel by oxidization of most of its carbon content in 
Bessemer converters or open hearth furnaces. Much 
of the gas from coke ovens of iron works is used as 
the fuel of large (at the Engine and Condenser 
Department of Allis-Chalmers, where I had a summer 
job in 1937, there was an eight foot bore by twelve 
foot stroke engine, for this purpose, on the drawing 
boards) internal combustion engines, which drive the 
blowers for the blast furnaces, etc. 

Petroleum is a mixture of hydrocarbons, which are 
separated, by distillation, into gases, including 
methane, butane, propane, and pentane, liquids, 
including naphtha, gasoline, kerosene, and fuel and 
lubricating oils, waxes, and asphalt. Asphalt, wax, 
and the "heavy" (viscous) oils owe their properties to 
the large size and complexity of their molecules, 
which, since the 1920s have been "cracked", at high 
temperature and pressure (which can be reduced by 



361 



using catalysts) to the smaller molecules of the more 
volatile compounds needed for internal combustion 
engines. 

Not long after I'd learned to read, I discovered "car 
cards", - advertising posters mounted in a row over 
the windows of a street car. One of these (advertising 
"Carbona", a dry cleaner) included a picture of a 
woman, with an article of clothing in one hand, flying 
through clouds, with the caption, "You can go twenty 
miles on a gallon of gasoline" as at least one 
automobile maker claimed at that time. My mother 
explained the point of the ad - that gasoline vapor 
could form an "explosive mixture" with air, but 
Carbona doesn't. That night, my father explained that 
Carbona is carbon tetrachloride which, though volatile, 
like gasoline (so it is useful for dry cleaning) but, 
unlike gasoline, it was not enflammable (gasoline 
forms explosive mixtures with air because it is highly 
enflammable). 

I accepted these explanations because they came from 
my parents, but didn't fully understand them at that 
age (between six and ten). The distinction between 
"enflammable" and "volatile" was unclear (as it still 
seems to be to some television news reporters). My 
father, sensing my perplexity , pointed out that 
"enflammable" meant "easily ignited to burst into 
flame" while "volatile" meant "easily evaporated". 
My view of the subject was obfuscated by the frequent 
spelling of "enflammable" as "inflammable", and the 
apparent interchangeability of the prefixes " in- " , 
"un-\ and "non-". (It would be fifteen years before 
I saw less ambiguous word "flammable" on a tank 
truck.) The meanings of such words as "explode", 
"explosion" "explosive", "detonate", and "detonation" 
and "detonable" were unclear to me then as they still 
seem to be to many. I am often, still, unsure what 
others mean by them, although I now think I know 
what I mean. 

Consideration of such matters from the various 
perspectives, acquired in the course of education, 
reading, conversations, and basement and backyard 
activities, clarified my views of fire, while presenting 
new perspectives, consideration from which led to 
other pictures, some of which were somewhat "fuzzy", 
leading to further research (literary, experimental, or 
analytical), a sequence which continues to this day. 

Following is an effort to summarize my picture of fire, 
as I now remember seeing it, in 1940 (which was two 
years after my graduation as a mechanical engineer): 



Most of the fires which I'd seen consisted of flames, 
which behaved as gases or colloidal suspensions (like 
smoke), and/or glowing coals. Clearly, most solid and 
liquid filets volatilize as part of the combustion process 
which showed me that more than the single step, 
implied by stoichiometric equations, such as Equations 
(1) and (2), are involved in the process. In the 
burning of gunpowder, expressed in equation (2): 

2KN0 3 + C + S -► C0 2 + SO 2 +2KO + N 2 

the potassium nitrate (KN0 3 ) molecules must 
dissociate to provide the oxygen atoms to oxidize the 
carbon and sulfur. Considered from the "interatomic" 
perspective, which has been discussed herein before, 
it seemed that the oxidation of hydrogen must follow 
the dissociation of the oxygen molecules, and probably 
those of hydrogen. I began to see fire, except where 
elemental carbon, as coal, coke, or charcoal, or in a 
similar form, is the fuel, is a multistage process. The 
decomposition, melting or sublimation, and/or 
evaporation, and condensation to the aerosol referred 
to as "smoke", precedes its further decomposition, 
dissociation, oxidation, and ionization, the effects of 
which are seen as the flames of familiar fuels (except 
carbon in its various forms). 

By 1940, I had acquired enough of the vocabulary of 
chemistry to understand that chemical processes and 
changes of state were either "exothermal" 
(characterized by the evolution of heat) or 
"endothermal" (characterized by the absorption of 
heat) and was aware that freezing, condensation, and 
the formation of most molecules by joining atoms, 
ions, or "free radicals" are exothermal, while melting, 
sublimation, boiling and other evaporation or 
vaporization, decomposition, ionization, and 
dissociation are endothermal. I came to recognize that 
"ignition" or "kindling", considered from the 
perspective outlined above, from which fire is seen as 
a multistage phenomenon, occurred only after 
sufficient heat had been transferred to a fuel element 
to result in the necessary succession of endothermal 
processes (which is unique for each fuel), and maintain 
the exothermal oxidation until the evolution of heat 
from the latter is sufficient to heat the "soot" (the 
particles of carbon which are products of the 
decomposition of the colloidally suspended droplets of 
hydrocarbons called "smoke") to the incandescence 
visible as "flame". Fire, in general, stabilizes when 
heat losses reach equilibrium with evolution of heat. 
When losses exceed evolution of heat, a fire "dies 
out". When the evolution of heat exceeds losses and 
continues to do so, the fire is self accelerating. A few 



362 



years later, the course of "literature research at the 
Naval Ordnance Laboratory, I was to read of such a 
self accelerating fire referred to as a " thermal 
explosion" (which will be discussed in more detail a 
few pages hence), but, in 1940, "explosion still meant 
to me, as it had since I learned the word, a "bang" 
and a flash and an impulse which could throw things 
around (As a kid, I had seen tin cans, propelled by 
firecrackers, fly higher than a house.)* and some tines 
broke them. Dictionaries (1,2) gave "detonation" as 
a synonym of "explosion", and an encyclopedia (15) 
stated that "detonation is a distinct phenomenon in 
which the chemical transformation is induced in every 
particle at (he same instant". Even then, I didn't 
believe that. I already saw fire as the Multistage 
phenomenon described above, of which each stage 
takes time. I had heard the combustion of an internal 
combustion engine referred to as "an explosion" (at the 
beginning of each power stroke), and the anti-knock 
property of Ethyl and other high octane gasoline 
ascribed to the fact that it was "slower burning" than 
regular. However, having considered the Otto cycle 
(which is employed in most automobiles), I was aware 
that even high octane gasoline burned fast enough that 
the reaction was complete before the piston moved 
enough that the volume change had to be considered in 
thermodynamic calculations, but "regular" , in a high 
compression engine, burned so fast that the resulting 
rapid pressure increase is propagated through the 
cylinder head to the air as a sound or "shock" wave. 
Waves had been familiar to since early childhood, 
when I saw them on water. My dad built a radio 
before I was ten and I soon learned to estimate where 
to set the dials from the wavelength of each station, 
published in the paper. I heard that radio waves were 
waves in "ether" but, before I found out what "ether" 
was, I read about the Michelson-Morley experiment, 
which showed that there was no such thing. As a 
teenager, I had built audio equipment, including 
amplifiers and recording equipment, in the course of 
which I acquired practical, empirical, and graphical 
perspectives of sound, which prepared me for 
acoustical and analytical perspectives of sound I was to 
learn in physics courses. In 1940, I was aware that 
both "explosion" and "detonation" are derived from 
Latin words describing sounds (those of clapping and 
thunder respectively), and that both referred to sudden 
fire (sudden enough that the pressure rise, due to the 
heat evolved, was propagated as a sound wave, from 
which I inferred that the reaction was completed) in a 
small fraction of a second (the maximum period of a 
sound wave), but my mental pictures of the processes 
were rather indistinct until, as a participant, at the 
Naval Ordnance Laboratory (N.O.L.), in the research 



and development of systems of which pyrotechnics and 
explosives are components. 

Although, in 1940, my impressions of explosion and 
detonation were not very clear, I could see, from a 
heuristic perspective of the Maxwell-Boltzmann kinetic 
theory of gases (as I understood it) that the combustion 
of an "explosive mixture" of gaseous fuel and air is 
the effect of random intermolecular collisions of 
sufficient magnitude to break (dissociate) molecules in 
into atoms, ions, free radicals. I saw that, in this 
state, hydrogen and carbon atoms and/or ions can bond 
to those of oxygen to form carbon dioxide and water, 
processes which are exothermal (evolving heat) - the 
kinetic energy of molecular motion) thus increasing the 
frequency of collisions of sufficient magnitude to 
initiate the sequence outlined above in the previously 
unreacted fuel/air mixture. I saw that this sequence 
should be expected to propagate at a velocity close to 
the average of those of the molecules (which is about 
that of sound). 

Consideration of the heuristic model, described above 
from a quantitative perspective, would have required 
statistical analysis beyond my abilities. Perhaps I 
could have clarified my view by consideration from 
acoustical, hydrodynamic, and/or fluid mechanical 
perspectives, to each of which I had been introduced, 
but I didn't make such an effort until, at N.O.L., I 
was engaged in explosive research, the course of 
which I learned that others, including Rankine (whose 
steam engine cycle I had learned of in thermodynamics 
courses) had done so in the nineteenth century. 

My earliest impressions of fire were effects of 
observations of and experience with such familiar fuels 
as paper wood, coal, charcoal, candle wax, and 
gasoline. I had became convinced that air is essential 
to fire. It didn't take long for experience with 
fireworks (which available, in late June and the first 
three days of July, in the 1920s, at grocery and drug 
stores, to any kid with a dime), to cast doubt on this 
conviction. My experience, at the "Universal 
Research Laboratory" (as a fourth grade classmate 
referred to his basement) in the preparation of black 
gunpowder, which burned quite vigorously in a rocket 
(which was lacking in aerodynamic stability and 
tumbled a few feet off the ground). I can't say, at this 
time (1995) whether, at that time (1926), I understood 
why gunpowder could barn without air, although other 
fuels couldn't. However, by 1940, I had learned that 
the burning of familiar fuels is oxidation, requiring the 
oxygen of air, but that gunpowder, as well as other 
explosives and pyrotechnics, burned without air 



363 



because they contained oxygen as a component of 
relatively unstable compounds, such as potassium 
nitrate. Thus, I saw, the burning of gunpowder is a 
multistage reaction (one stage of which is the 
decomposition of the nitrate) which, in fuses, is too 
slow to be called H an explosion". 

In 1940, World War II had become the "Battle of 
Britain", Japan was rearming the invading Asiatic 
neighbors, and the U.S. armed services were engaged 
in an effort to regain and surpass the preparedness lost 
as the result of the pacifism and disarmament 
following World War I. Pursuant to this effort, the 
U.S. Naval Ordnance Laboratory (NOL) recruited 
over 1700 scientists and engineers, of which I was 
one. 

FIRE, AS I SAW IT AT NOL (EAU) 

Arriving at NOL, in early 1941, I was given new 
perspectives, particularly of fire, too frequently to 
retrace after fifty-odd years. My first assignment, at 
NOL, was to the Propellant and Pyrotechnics Group of 
the Experimental Ammunition Unit (EAU), where I 
worked with an "ordnance man", directed by the 
pyrotechnics specialist on the adaptation of display 
fireworks for use as signals (such as "Submarine 
Emergency Identification Signals"). My principal 
assignment, in that group was design and drafting, but 
I spent some time in the laboratory, where we did 
some preparation, fabrication, and testing of 
pyrotechnic mixtures, systems, subsystems, 
components, etc. Our approach, in such adaptation, 
was generally that of "trial and error" or, to dignify it, 
"Edisonian research". Based on the view, from the 
"practical" or "common sense" perspective, that 
mixtures and practices which had been used, with 
success, by others, were most likely to serve our 
similar purposes, we usually followed recipes, some 
dating back to those of medieval alchemists and 
ancient Chinese artisans. On occasion, we tried to 
improve mixtures by application of stoichiometry, but 
experience tended to confirm the above mentioned 
view, from the "practical" perspective. It was 
apparent that more than stoichiometry was involved. 
We noted that the behavior and properties of 
pyrotechnic mixtures were affected by the granulation 
of their ingredients as well as the densities to which 
they were loaded. Of the properties so affected, 
sensitivity, which includes their susceptibility to 
initiation by the stimuli available for this purpose when 
intended, as well as that to initiation by accidentally 
applied stimuli (and resulting hazards), is of primary 



concern to all involved in the manufacture and 
application of pyrotechnics and explosives. 

Although, from instruction, conversation, and "hands 
on" experience, I had acquired some "eyewitness", 
"common sense", "practical", and "empirical" 
perspectives of such matters, I felt a lack of applicable 
"chemical", and "scholarly" perspectives. The EAU 
had no official library, but I had noticed a number of 
books on ordnance, explosives, and pyrotechnics on 
desks and shelves, which I borrowed and read. One, 
which caught my eye was The Chemistry of Powder 
and Explosives" (4), by Tenney L. Davis (who had 
been the lecturer of my second semester freshman 
chemistry course). The book was intended to be a 
textbook for a graduate course on the subject, the 
book includes ( in the 1941 edition) chapters on 
PROPERTIES OF EXPLOSIVES, BLACK POWDER, 
PYROTECHNICS, AND AROMATIC NITRO 
COMPOUNDS. It starts by defining an "explosive as: 

"a material, either a pure of single substance or a 
mixture of substances which is capable of producing 
an explosion by its own energy". 

"It seems unnecessary to define an explosion, for 
everyone knows what it is. ~ a loud noise and the 
sudden going away of things from where they had 
been --". 

As I read it, I accepted it, but soon began to recognize 
that, as with many words, although everyone knows 
what "explosion" means, it doesn't necessarily mean 
the same to everyone. Recently, a steam automobile 
enthusiast told me that "Explosion is the most 
inefficient form of combustion", implying, I support 
that a steam automobile should be more efficient than 
one with an internal combustion engine. Do those 
who talk about "the population explosion" mean "a 
loud noise - etc." or "the most inefficient form of 
combustion"? 

Davis, in the first chapter on PROPERTIES OF 
EXPLOSIVES, points out that, although an explosive 
can produce an explosion by its own energy, it can 
liberate this energy without exploding (as black 
powder does in a fuse). Here, he injected an 
explanation of the difference between a "fuse" which 
is a device for communicating fire" and a "fuze", 
which is a device for initiating explosion (usually 
"detonation") of the "bursting charges" of shells, 
bombs, mines, grenades, etc.". A section of the first 
chapter on PROPERTIES OF EXPLOSIVES headed, 
Classification of explosives includes paragraphs 
(condensed below) on: 



364 



I. Propellants, or "low explosives" which (in their 
usual application) burn but do not explode, and 
function by producing gas which produces an 
explosion (by busting its container, such as the paper 
tube of a Chinese firecracker or the metal case of a 
bomb). Examples: black powder, smokeless powder. 

II. Primary Explosives or "initiators" , which explode 
or detonate when heated or subjected to shock. 
Examples: mercury fulminate, lead azide, lead salts of 
picric acid, etc. 

III. High Explosives, which detonate under the 
influence of the shock of the explosion of a suitable 
primary. Examples: dynamite, TNT, tetryl, picric 
acid, etc. 

It is pointed out that these classes overlap because the 
behavior of explosives is determined by the nature of 
the stimuli to which they are subjected and by the 
manner in which they are used. Nitrocellulose, 
"colloided" such smokeless powder, is a propellant, as 
compressed guncotton, is a powerful high explosive, 
and, as lower density guncotton, has been used as the 
"flash charge" of electric detonators, TNT, 
nitroglycerine, and other high explosives have been 
ingredients of smokeless powder. Mercury fulminate 
can be "dead pressed" so that it loses its power to 
detonate from flame. 

A review of the foregoing two pages has led me to 
recognize the need for an explanation. Although the 
discussion FIRE AS I SAW IT AT NOL (EAU) is, as 
implied based on my memories after fifty-some years, 
the review of Davis's book (4), is a reflection of 
current "browsing" through a copy at hand. The 
quotations enclosed in quotation marks, including the 
following, are direct copies. 

"Propagation of Explosion" 

"When black powder burns the first portion to receive 
the fire undergoes a chemical reaction which results in 
the production of hot gas. The gas, tending to expand 
in all directions from the place where it was produced, 
warms the next portion of black powder to the kindling 
temperature. This then takes fire and burns with the 
production of more hot gas which raises the 
temperature of the next adjacent material. If the black 
powder is confined, the pressure rises, and the heat, 
since it cannot escape, is communicated more rapidly 
through the mass. Further, the gas- and heat- 
producing reaction, like any other chemical reaction, 
doubles its rate for every 10° (approximate) rise of 



temperature. In a confined space the combustion 
becomes extremely rapid, but it is believed to be 
combustion in the sense that it is a phenomenon 
dependent upon the transmission of heat. " 

"The explosion of a primary explosive or of a high 
explosive, on the other hand, is believed to be a 
phenomenon which is dependent upon the transmission 
of pressure or, perhaps more properly, upon the 
transmission of shock. Fire, friction or shock, acting 
upon, say, fulminate, in the first instance cause it to 
undergo a rapid chemical transformation which 
produces hot gas and the transformation is so rapid 
that the advancing front of the mass of hot gas 
amounts to a wave of pressure capable of initiating by 
its shock the explosion of the next portion of 
fulminate, and so on, the explosion advancing through 
the mass with incredible quickness. In a standard No. 
6 blasting cap, the explosion proceeds with a velocity 
of about 3500 meters per second. " 

"If a sufficient quantity of fulminate is exploded in 
contact with trinitrotoluene, the shock induces the 
trinitrotoluene to explode, producing a shock adequate 
to initiate the explosion of a further portion. The 
explosive wave traverses the trinitrotoluene with a 
velocity which is actually greater than the velocity of 
the initiating wave in the fulminate. Because this sort 
of thing happens, the application of the principle of the 
booster is possible. If the quantity of fulminate is not 
sufficient, the trinitrotoluene either does not detonate 
at all or detonates incompletely and only part way into 
its mass. For every high explosive there is a 
minimum quantity of each primary explosive which is 
needed to secure its certain and complete detonation. 
The best initiator for one high explosive is not 
necessarily best initiator for another. A high explosive 
is generally its own best initiator unless it happens to 
be used under conditions in which it is exploding with 
its maximum "velocity of detonation". 

The section of Davis's book quoted above presents 
views of explosion and detonation from "common 
sense" and "empirical" perspectives. I am sure Davis 
was aware of "theoretical", "analytical" 
"hydrodynamic", etc. views which had been expressed 
in the previous century or more, but confined his 
description to views from perspectives with which, he 
felt confident, his students and other readers of his 
book would be familiar. References to the "kindling 
temperature" and the "doubling of reaction rate for 
every 10° rise of temperature" are evidence of an 
"empirical" perspective based on standardized tests 



365 



similar to those described a few pages on in the first 
chapter of the book (4). 

The section of Propagation of Explosion is followed 
by a section on Detonating Cord which is also 
referred to as "cordeau" (after the French "cordeau 
detonant") and "Primacord" (a trademark of the 
Ensign Bickford Company), a narrow tube filled with 
high explosives whose principal use in blasting is the 
simultaneous (or in close sequence) initiation of 
detonation of two or more explosive charges. It can 
also be used to fell small trees as had been 
demonstrated in an ROTC class I had attended a few 
years before. 

The first chapter of "The Chemistry of Powder and 
Explosives" (4), entitled PROPERTIES OF 
EXPLOSIVES also includes descriptions of 
experiments, which can be performed in a college 
chemistry laboratory to demonstrate some of these 
properties, as well as standardized tests for them, and 
cites U.S. War Department Technical Manual TM 
2900 "Military Explosives" (25) and a number of 
Bulletins and Technical Papers of the U.S. Bureau of 
Mines in which such standardized tests are described 
in more detail. Properties for which tests are 
described in this chapter include Velocity of 
Detonation Sensitivity (including "explosion" , 
"ignition", or "kindling" temperature and impact 
sensitivity) and Tests of Power and Brisance. 

The section on Velocity of Detonation begins with a 
paragraph in which the subject is considered from the 
"empirical" perspective of the time (ca 1940), which 
no longer seems relevant. It mentions detonation 
velocity measurements by Berthelot and Vielle, who 
used a "Boulenge* chronograph" which is not 
described, except for the mention that its (lack of) 
precision was such that they were obliged to employ 
long columns of explosives". It goes on to say, "The 
Mettegang recorder, now commonly used, is an 
instrument of much greater precision and makes it 
possible to work with much shorter charges". The 
Mettegang recorder is described as an instrument 
whereby time is determined as proportional to the 
distance (measured with a micrometer microscope) 
between marks made on a rapidly moving smoked 
metal surface. 

Also mentioned is the "Dautriche method" in which 
the detonation velocity of an explosive being 
investigated is compared with that of detonating cord, 
which can be determined using relatively imprecise 
timers with long lengths of the cord. 



The use of high speed photography and cathode ray 
oscillographs, in measurement of detonation velocity, 
are mentioned as recent developments. 

The section on Sensitivity Tests includes descriptions 
of "impact" or "drop" tests in which (typically) the 
height is determined which will result in the explosion 
of a sample of the explosive being investigated 
contained in a hole in a block of steel, when a two 
kilogram weight is dropped on a plunger in contact 
with the explosive sample. 

Also described in a test of "temperature of ignition" or 
"explosion temperature" in which a blasting cap cup 
containing a sample of an explosive is thrust into 
Woods metal which has been heated to a known 
temperature. The procedure is repeated, with varying 
temperatures, until the temperature is determined at 
which the sample explodes or is ignited within five 
seconds. In another test which is also described, the 
blasting cap cup containing the explosive being tested 
is dunked in Wood's metal at 100°C and the 
temperature is raised at a steady rate until the sample 
ignites or explodes. The temperature at which this 
occurs is considered to be the "ignition" or "explosion 
temperature". When the temperature is raised more 
rapidly, the inflammation occurs at a higher 
temperature". (As is illustrated in an accompanying 
table). 

In the section on Tests of Power and Brisance a 

definition of neither "power" nor "brisance" is 
included. It seems that, as with "explosion", Davis 
assumed that it was unnecessary to define "power" 
because everyone knew what it meant, while, in my 
view, as with "explosion", although everyone knew 
what "power" meant, it didn't mean the same to 
everyone. (Dictionaries I've consulted (1,2,16) 
include from eight to seventeen definitions) . 
"Brisance" on the other hand, is not listed in a 1939 
unabridged dictionary (2). More recent dictionaries 
(1,16) define it as "The sudden release of energy by a 
explosion", or something similar. Although "brisance" 
wasn't defined in Davis (4) or the dictionaries I 
consulted (2), it didn't take long for me to grasp its 
meaning, whether from conversation, recognition 
(having studied French in high school) that it was 
derived from "briser"-to break, or from descriptions, 
in Davis (4) and references cited therein, of tests of 
power and brisance (each of which rated an explosive 
in terms of the measurement of the deformation or 
other change of a solid specimen which had been 
exposed to its action. 



366 



The opening paragraph of the section on Tests of 
Power and Brisance mentions a "manometric bomb" 
as a means of measuring the energy liberated in an 
explosion, but goes on to point out that the 
effectiveness of an explosion depends upon the rate at 
which the energy is liberated ("Power" as understood 
by physicists and engineers). The high pressures 
developed by explosions (which reflect this rate) were 
first measured by the Rodman gauge, in which, 
according to Davis (4), the pressure caused a hardened 
steel knife to penetrate into a disc of soft copper. The 
depth of the penetration was taken as a measure of this 
pressure. Davis also mentions "crusher Gauges" in 
which copper cylinders are crushed between steel 
pistons, piezoelectric gauges, the "Trauzl lead block 
test", in which the enlargement of a hole is taken as a 
measure of "power" or "brisance", the "small lead 
block test" of the Bureau of Mines, in which the 
compression of the block is used a such a measure, the 
"lead plate test of detonators" in which the diameter of 
the hole punched through the plate by a detonator is a 
measure of its output. Similar tests with aluminum 
plates are also mentioned. 

Explosive research was not, in 1941, among the 
missions of the NOL, but tests similar to those of 
PROPERTIES OF EXPLOSIVES mentioned and 
described by Davis (4) in the chapter so titled, which 
is reviewed above, were performed by ordnance men 
assigned to those Fuze Group of the EAU (whose 
laboratory, we of the Propellant and Pyrotechnics 
Group shared, so I could watch now and then) 
pursuant to the development of fuze explosive trains 
(an activity in which I was to become engaged in a 
few months, so that I acquired "hands on" experience 
with such tests). Meanwhile, however, my job was to 
help develop pyrotechnic signals as a member of the 
Propellants and Pyrotechnics Group. 

The adaptations of display fireworks, which were the 
objects of our efforts, were, for the most part, charges 
of mixtures of fuels such as charcoal, sulfur, 
magnesium, sugar, aluminum, iron filings, and sodium 
oxalate with oxidants, such as potassium, sodium, 
barium, and strontium nitrates, and potassium chlorate, 
and perchlorate, which were usually initiated by the 
flames from fuses, such as Bickford safety fuse. 
Sensitivity to such initiation was characterized in terms 
of the maximum gap between the end of the fuse and 
the surface of the pyrotechnic charge at which the 
charge was ignited. It seemed to me that the 
relationship between particle size, compaction, and 
sensitivity was similar to the relationship between 
analogous features of the twigs, shavings, or other 



"tinder" and ignitability, which I had noted when 
passing the Boy Scout second class fir building test in 
1927. As I saw it then (in 1941) ignition required that 
some of the fuel be raised to its "ignition 
temperature", a concept in which I believed (with 
some reservations) at that time. This seemed to be 
more easily accomplished with small elements of fuel 
in not too intimate contact with others. 

We used black gunpowder for various purposes, 
including augmentation of primer output to ignite flair 
compositions. 

In this connection, I learned that "cannon powder" was 
the coarsest of those we used because black powder as 
well as other propellants, burns at the surface of the 
grains so that the coarser grains burn longer to 
maintain the generation of gas over the longer time 
that a cannon projectile spends in the barrel. I later 
read (in Davis (4)) that the "grains" of smokeless 
powder for large guns are perforated to provide an 
increasing surface area as the burning progresses and 
the holes enlarge so that the gas emission rate (which 
is proportional to the area of the burning surface) 
keeps pace with an accelerating projectile. 

After a year or so with the Propellant and Pyrotechnic 
Group, I was transferred to the Research Group of the 
EAU, (supervised by Harry H. Moore) which, as I 
remember, took on all tasks assigned to the EAU, 
which were not obviously within the purview of the 
Propellant and Pyrotechnic Group or the Fuze Group. 
The Research Group also was responsible for the 
design and development of some hydrostatic bomb 
fuzes for antisubmarine warfare (probably because, at 
some earlier date, the Fuze Group had been too busy 
with other projects), an area in which I found myself 
shortly after my transfer to the Research Group. My 
attempt to design an improved hydrostatic fuze led to 
the assignment of an attempt to establish fuze 
explosive train design criteria. Before asking Mr. 
Moore for fuze explosive train design criteria, I asked 
members of the Fuze Group, who had designing fuzes 
for several years. Their approach as I understood it, 
was that of adapting a successful explosive train to a 
new fuze, showing their proposed design to their boss, 
Mr. Ray Graumann. If he approved, They'd have 
some made and test them. Although I was to find out, 
some years later, that this was a reasonably sound 
approach (since Graumann was a nationally recognized 
authority on fuze design), it didn't satisfy me, or Mr. 
Moore, at the time, all of which led to the assignment 
mentioned above. From the members of the Fuze 
Group, I learned that the word "fuze" was derived 



367 



from "fuse", (I had read in Davis (4)) the distinction 
between the terms, which is quoted hereinbefore and, 
they said, "fuze" had been defined "by act of 
Congress" (as has been quoted, from Davis (4) 
herein). 

About the only "fuze explosive train design criterion" 
I learned from them was the requirement for "of-line 
safety" to preclude the transmission of detonation from 
a detonator to the bursting charge until after a fuze had 
"armed". 

Application if this criterion requires a definition or 
standard of "detonation". The EAU Fuze Group used 
the visible damage to the metal parts which had held 
the explosive charge for this purpose, which, in turn, 
requires a practiced eye. Members of the Fuze Group 
could, at a glance, recognize evidence of detonation, 
and even characterize the detonation as "high order" 
or "low order". Some even identified certain damage 
as evidence of "high intensity low order. 

I was one of over 1700 engineers and scientists 
recruited by NOL in 1939, '40, and '41 to meet the 
increasing demands of the preparedness effort in 
response to the rising hostilities in europe and the 
perceived probability that the US would be involved. 
The new employees, representing a wide range of 
technical professions and coming from all over the 
country, saw things from many perspectives, to which 
we introduced on another. Few of us found ourselves 
in a position to apply our previous (specific) 
experience, but most could apply our general technical 
backgrounds. Lunch time conversations covered a 
wide range of subjects, mostly more or less technical. 
Sometime, in 1 94 1 , I remember hearing that 
"Detonation is a special kind of fire. " Somewhere else 
I heard detonation referred to as a "chain reaction". 
I felt the need for a more meaningful (to me 
definition. It seemed to me that such a definition 
should be more "scientific" than "a special kind of 
fire", "a chain reaction", or "explosion", (as it is still 
defined in dictionaries (1, 2, 13)). I didn't believe 
that "detonation is a distinct phenomenon in which the 
chemical transformation is induced in every particle of 
the mass of explosive at the same instant" , as defined 
in an encyclopedia (14). Davis' (4) description of 
"Propagation of Explosion" which is quoted herein (a 
few pages back) was more meaningful to me. Davis 
(4) included a heuristic description of what he referred 
to as the "propagation of explosion" in black powder 
(which, now, seems applicable to other "energetic 
materials" which have gaseous reaction products). 
Based on impressions I'd gained, from conversations 



with other EAU employees, that the burning rate of 
black powder (and other propellants and explosives) is 
proportional to pressure, I determined by a simple 
integration that the pressure, and hence the burning 
rate of a confined charge of black powder (or other 
propellant or explosive) must increase exponentially 
with the passage of time until its container, such as a 
bomb case or the paper tube of a firecracker, bursts. 

What Davis meant by "powder and explosives" is what 
aerospace engineers seem to mean by "pyrotechnics" 
and what others mean by "energetic materials." 

Davis' (4) description provided a perspective of fire 
(or combustion) including explosion and detonation, 
which, as Davis pointed out in the paragraph quoted a 
few pages back, are forms of combustion in that they 
are dependent upon the transmission of heat. From 
this explanation, combined with the thermodynamics I 
had learned in engineering school, I began to see 
detonation as "The rapid and violent form of 
combustion in which energy transfer is by mass flow 
in strong compression waves. " Although these words 
are taken from the opening sentence of "Detonation" 
by Weldon C. Ficket and William C. Davis (16), 
(which was published in 1979) they express, more 
adequately than any that come to mind in 1994, the 
view I received in 1942 from the book (4) Tenney L. 
Davis had published the previous year. This view is 
quite similar to that held today by those who have 
concerned themselves with detonation, but it is not 
sufficiently quantitative to serve as a basis for fuze 
explosive train design criteria. The need for further 
research was quite apparent. 

My assignments, in the Research Group of the EAU, 
besides the attempt to establish fuze explosive train 
design criteria, were various, including the adaptation 
of a motion picture camera (designed for use in the 
motion picture industry) for "slow motion" recording 
(from aircraft) of surface effects of underwater 
explosions, adjusting designs of coil springs, design of 
loading presses, fuze test gear, blast gages, and mine 
detonators (to replace those which were adaptations of 
Nobel's original (1864) blasting cap, (and still in use 
in the 1940s) with an adaptation of modern (as of the 
1940s) commercial blasting caps. When NOL started 
using "time sheets" to generate records of the 
distribution of effort among projects, I found myself 
trying to remember how many hours I had spent, 
during each past week on each of 42 projects. 

Research pursuant to the establishment of fuze 
explosive train design criteria, at this stage, literature 



368 



research, was fit in between other, more urgent 
efforts. I started with books on desks and shelves in 
the EAU, including Davis (4), in which I reviewed the 
parts which have been quoted hereinbefore, and 
finished reading the 1941 edition, and when it turned 
up, the 1943 edition. A couple of books on ordnance 
and explosives, the titles and authors of which I have 
forgotten, provided a historical perspective, but didn't 
have much which seemed relevant to fuze design. 

The chapters of Davis (4) on BLACK POWDER, 
PYROTECHNICS, and AROMATIC NITRO 
COMPOUNDS broadened and sharpened my historical 
and chemical perspectives of fire (particularly in the 
forms of explosion and detonation). The use of 
"Greek fire", a mixture of saltpeter (potassium nitrate) 
with combustible substance as an incendiary in naval 
warfare is cited as a progenitor to the discovery that a 
mixture of potassium nitrate, charcoal, and sulfur is 
capable of doing useful mechanical work, and the 
invention of the gun. The chapter traces the evolution 
of black powder for blasting as well as use as a 
propellant and includes tables of the proportions of 
charcoal, sulfur, and saltpeter as described by Marcus 
Graecus in the 8th century, Roger Bacon in the 13th 
and others in the 14th, 16th, 17th, and 18th centuries, 
as well as the stoichiometric equation (2): 

2KN0 3 + C+S -► C0 2 +S0 2 +2KO 

for the burning of black powder, and was aware that 
black powder burns without air because the thermal 
decomposition of the nitrate releases enough oxygen to 
oxidize the charcoal, sulfur, and potassium, while the 
nitrogen forms its own (N2) molecules. 

Davis' (4) chapter on PYROTECHNICS, by which he 
meant display and amusement fireworks (including 
such noisemakers as firecrackers, flares, signals, etc.), 
traces the development of such items from ancient 
oriental origin to the fireworks we played with as kids 
on the Fourth of July, and the signals and illuminating 
flares used by the military, and warning flares used by 
railroads and highway repair crews. This chapter had 
been of specific interest to me as a member of the 
Propellant and Pyrotechnics Group, providing 
"chemical" and "historical" perspectives of materials 
with which we were working. Included were tables of 
compositions for colored lights, flares, fuzes, smokes, 
marine signals, parade torches, whistles, lances, 
rockets, Roman candles, stars, fountains, pinwheels, 
mines, comets, meteors, torpedoes, flash cracker 
compositions, sparklers, serpents, snakes, etc. (Many 
of the foregoing terms were, apparently, used in the 



sense which had been familiar to me as a kid, 
shopping in late June and early July in preparation for 
celebration of the Fourth, rather that in the senses, in 
which they are used in connection with biology, 
ordnance, astronomy, or defined in most dictionaries). 
Consideration of these formulas led to recognition that 
nitrates other than that of potassium, as well as 
chlorates perchlorates, etc., have been used as solid 
providers of oxygen, and increased my awareness that 
firelight, other than the (black body radiation) of 
familiar yellow flames, was spectral emission of some 
elements (such as strontium, which emits red light) in 
the plasma state and that colored smoke in a colloidal 
suspension of dye in air. 

Davis (4) chapter on AROMATIC NITRO 
COMPOUNDS, after an introductory paragraph on 
their usefulness, in which it is stated that they were 
(when the book was published - in 1941) "the most 
important class of military high explosives" warns of 
their toxicity and the fact that they can be absorbed by 
the skin. A paragraph on the chemistry of these 
compounds was probably elementary to the graduate 
students for which the book was written, but for a 
mechanical engineer, such as I, it was a bit cryptic. 
I'm not sure, in the 1990s, when I learned that 
"aromatic compounds" include "benzene rings" "a 
structural arrangement — marked by six carbon atoms 
lined by alternate single and double bonds" (2a), and 
that a "nitro group" (NO^ the oxygen atoms are 
bonded to the nitrogen atom which in an aromatic nitro 
compound is, in turn bonded to one of the carbon 
atoms. 

Considering the chemistry of aromatic nitro 
compounds as discussed by Davis (4), and quoted 
above, from the basic chemical perspective I had 
acquired in chemistry courses, I saw that, like 
gunpowder and other pyrotechnics, they contained 
oxygen sufficient to oxidize at least some of the carbon 
and hydrogen they contained, but that the oxygen 
would be available only after decomposition of the 
nitro groups (or, in the case of black powder, of the 
potassium nitrate). I can't say, now, when I looked up 
the heats of formation of the compounds involved, but, 
when I did, I verified that the reactions were 
exothermal. These relationships were, of course, 
obvious to the graduate students in chemistry for 
which Davis' book (4) was intended, as was the fact 
that, at the anticipated temperatures of the reaction 
products (carbon dioxide and water) they were in a 
gaseous state, so that the product of their pressure and 
volume was many times those of unreacted solid or 
liquid unreacted explosives and propellants, as well as 



369 



those of gases or explosive mixtures of gases or 
gases and sols (colloidal suspensions) This, their 
combustion was considered to be an explosion or 
detonation if it was fast enough - which, in turn, turns 
on definitions of explosion and detonation . 

EXPLOSION AND DETONATION 

Explosion and detonation are among the many 
words which, like fire and pyrotechnics have a 
number of meanings. According to some dictionaries 
(1,2), they are synonymous, although I doubt that any 
participant in this workshop considers them to be. 
Each of the words has a connotation of sudden 
expansion , and each is derived from a Latin word 
relating to the sound usually associated with it. 
Detonation (some meanings of which will be 
discussed below) is derived from the Latin for thunder 
while explosion (like applause ) derives from the 
Latin for clap , but it seems to be mostly commonly 
used in the sense of bursting of a container such as 
a boiler or a bomb, but it often means sudden burning 
as in a dust explosion or that of another explosive 
mixture of gaseous or finely divided fuel with air. 

As observed by chemists and stated in the rule of 
thumb which has been mentioned in the quotation of 
Davis' (4) section on Propagation of Explosion that 
any — chemical reaction, doubles its rate with each 
10°C (approximate) rise of temperature, so that 
exothermal reactions (in which heat is evolved) tend 
to be self accelerating. If, as is usual, their reaction 
products include one or more gases, the pressure rises, 
if they are confined, until the container bursts. Since 
surface or grain burning rates of propellants, 
explosives and explosive mixtures increase with 
increasing pressure (17), such burning is self 
accelerating and the pressure and rate increase, 
exponentially with time. Either the bursting of a 
container or self accelerating process is an explosion 
in one or another of the senses mentioned above. 
(Self accelerating fire is referred to as thermal 
explosion ). This association has led to the use of 
explosion to mean any self accelerating process, such 
as in population explosion (which seems sudden 
only from such perspectives as historical or 
geological ). 

In lunchtime conversations, I had hear detonation 
defined in several sets of terms. Dictionaries and 
encyclopedias included definitions and descriptions 
which didn't satisfy me. Davis' (4), discussion (which 
has been quoted herein) gave me the clearest (though 
still somewhat fuzzy) view of the process I had 



acquired until my literature search in the areas of 
explosives and detonation extended beyond the shelves 
and desks of the EAU. Davis' (4), in the section on 
Propagation of Explosion ,, which has been quoted 
herein, made the transition to the discussion of 
detonation with the statement that The explosion of a 
primary explosive or of a high explosive is believed to 
be a phenomenon which is dependent upon pressure 
or, perhaps more properly, on transmission of shock 
. The propagation of detonation mercury fulminate is 
described as a reaction which produces hot gas and is 
so rapid that the advancing front of the mass of hot 
gas amounts to a pressure wave capable of initiating 
by its shock the next portion of fulminate . The high 
velocities of propagation of detonation (3500 meters 
per second in the fulminate of a blasting cap and much 
higher in TNT) are mentioned. Although Davis (4) 
didn't consider shock or detonation waves from 
physical, thermodynamic, or hydrodynamic 
perspectives, his discussion left me with the impression 
that consideration from such perspectives would be 
appropriate, an impression which was to be verified 
when I started reading OSRD reports. 

As I recall, the other books on the desks and shelves 
of the EAU, which included a couple of books on 
ordnance, one, by a hobbyist, on fireworks, and the 
Dupont Blasters Handbook , considered their subjects 
from practical , empirical , and historical 
perspectives, omitting discussion of chemical or 
physical aspects of their subjects. 

On the NOL library, I found the War Department 
(which, after a decade or so, was to become the 
Department of the Army) and Bureau of Mines 
documents which had been cited by Davis (4) and 
others from these sources and from Picatinny Arsenal, 
Frankford Arsenal, the Ballistics Research Laboratory 
(BRL) at Aberdeen, MD., the Naval Powder Factory 
at Indian Head, Md., the Franklin Institute, the Bureau 
of Standards, and the British Ministry of Supply and 
Ordnance Board. I found these documents 
informative, interesting, and (some of them) quite 
pertinent to my objective of establishing fuze explosive 
train design criteria, but none contained much to 
clarify my views of explosion or detonation except for 
a few OSRD reports. 



370 



REFERENCES 

(1) Stein, Jess, (editor), The Random House 
Dictionary of the English Language , Random House, 
Inc. 1966-1967. 

(2) Webster's Twentieth Century Dictionary of the 
English Language (unabridged) . Publisher's Guild, 
Inc., New York, N.Y., 1939. 

(2V£) McCrone, John C, The Ape that Spoke . 
William Morrow and Company, Inc., New York, 
N.Y., 1991. 

(3) Calvin, William H., The Ascent of the Mind. Ice 
A ge Climates and the Evolution of Intelligence . 
Bantam Books, New York, Toronto, London, Sydney, 
Auckland, 1990. 

(4) Davis, Tenney L., The Chemistry of Powder and 
Explosives . John Wiley & Sons, Inc., New York, 
1941, 1943. 

(5) Asimov, Isaac, Asimov's Biographical 
Encyclopedia of Science and Technology . Doubleday 
& Company, Inc., Garden City, N.Y., 197. 

(6) Gamow, George, One Two Three... Infinity . The 
Viking Press, New York, 1947. 

(6V4) Bradbury, Fahrenheit 451 . Simon and Sinister, 
New York. 

(7) Wells, Robert W., Fire and Ice. Two Deadly 
Wisconsin Disasters. Fire at Peshtigo . North word, 
Madison, WI, 1968. 

(8) Gamow, George, Thirty Years That Shook 
Physics . Doubleday & Co. (Science Study Series). 
New York, N.Y. 1966. 

(9) Langdon-Davies, John, Inside the Atom . Harper & 
Brothers, New York and London, 1933. 

(10) Weast, Robert C. (editor), and Astle, Melvin J. 
(associate editor), CRC Handbook of Chemistry and 
Physics. 63rd Edition . CRC Press, Inc., Boca Raton, 
FL. 1982-1983. 



(12) Chapman, D.L., "On the Rate of Explosion of 
Gases", Philosophical Magazine (5), V. 47, p. 90, 
1999. 

(13) Jouguet, E., 4echanique des Rxploszifs" Paris, O. 
Doin et Fils, Editeurs, 1817. 

(14) The Grolier Society, Grolier Encyclopedia) . Vol. 
4, 1931-1951. 

(15) Brady, George S., and Clauser, Henry R., 
Materials Handbook . McGraw-Hill Book Co., New 
York, etc., etc. 1977. 

(16) Soukhanov, Anne M., (Executive Editor), The 
American Heritage Dictionary of the English 
Language. Houghton-Mifflin . Boston, New York, 
Londom, 1972. 

(17) Ficket, Wildon, and Davis, William C. 
Detonation , University of California Press, Berkeley, 
Los Angeles, London, 1979. 

(18) Smithsonian Tables . 

(19) Henkin, T. "Determination of Explosion 
Temperatures", OSRD 1986, November 1943. (Later 
published with McGill, R. , in the Journal of 
Engineering and Industrial Chemistry 44, 1391, 1952. 

(20) Kabik, I., Rosenthal, L. A., and Solem, A.D., 
The Response of Electroexplosive Devices to 
Transient Electrical Pulses" 3rd Electrical Initiator 
Symposium, Paper #18, 1960. 



[Editor's Note: Time and space precludes completion 
of Mr, Stresau's paper. He does plan to publish the 
entire story as a Stresau Laboratories, Inc. report at 
a later date. Interested readers may contact him at: 

Stresau Company 

W7882 Stresau Lane 

Spooner, WI 54801 J 



(11) Eschbach, Ovid W., Handbook of Engineering 
Fundamentals. Third Edition . Wiley, New York. 



371 



APPENDIX - List of Participants 



Aerojet Propulsion Division Hansen, Jeff 

Aerospace Corporation, The Gageby, James 

Aerospace Corporation, The Goldstein, Selma 

Aerospace Corporation, The Wong, T. Eric 

American Safety Flight Systems, Inc. . . . Ziegler, Ron 

Analex Corporation Smith, Floyd 

Analex Corporation Steffes, Paul 

Attenuation Technology, Inc Dow, Robert L. 

Boeing Defense & Space Group . . Robinson, Steven P. 

Conax Florida Corporation Nowakowski, Don 

Consultant Folsom, Mark 

Consultant Gans, Werner A. 

Consultant O'Barr, Gerald L. 

EG&G Mound Applied Tech. . . . Beckman, Thomas M. 

EG&G Mound Applied Tech Kramer, Daniel 

EG&G Mound Applied Tech Munger, Alan C. 

EG&G Mound Applied Tech Spangler, Ed 

Energetic Materials Technology .... Ostrowski, Peter 

Ensign Bickford Aerospace Co Graham, John A. 

Ensign Bickford Aerospace Co Renfro, Steven L. 

Ensign Bickford Aerospace Co Rhea, Arthur D. 

Franklin Applied Physics Thompson, Ramie 

Geo-Centers, Inc Landry, Murphy J. 

Halliburton Energy Services 

Explosive Products Center . . . Barker, James 
Halliburton Energy Services 

Explosive Products Center . . . Motley, Jerry 

Hercules Aerospace Cole, David A. 

Hercules Inc McAllister, Pat V. 

Hi-Shear Technology Corp Novotny, Don 

Hi-Shear Technology Corp Webster, Richard 

Inland Fisher Guide Div. of GM Wirrig, Steven T. 

John Hopkins University - CPIA Filliben, Jeff 

Los Alamos National Laboratory .... Kennedy, James 

Martin Marietta Specialty Components . . . Hinkle, Lane 

Martin Marietta Astronautics Wood, Lance 

McDonnell Douglas Tierney, Michael 

McDonnell Douglas Aerospace Whalley, Ian 

McDonnell Douglas Space Sys. . . Parenzan, James E. 

Morton International Inc Hansen, David 

Morton International Inc Richardson, Bill 

NASA Headquarters Schulze, Norman R. 

NASA Goddard Space Flight Center . . Bajpayee, Jaya 

NASA Johnson Space Center Hoffman, William 

NASA Kennedy Space Center Rayburn, Larry 

NASA Langley Research Center Bement, Larry 

NASA Lewis Research Center Seeholzer, Tom 

NASA Stennis Space Center .... St. Cyr, William W. 
Naval Surface Warfare Center 

Indian Head Division Blachowski, Tom 

Naval Surface Warfare Center 

Indian Head Division Hinds, Jeffery L. 

Naval Surface Warfare Center 

Indian Head Division Krivitsky, Damn 

Naval Surface Warfare Center 

Indian Head Division Martin, Steve 

Naval Surface Warfare Center 

Crane Division Schlamp, Jan N. 



A- 1 



Olin Rocket Research Co Mo ran, Joe 

Olin Rocket Research Co Watson, Bruce 

Pacific Scientific Corp Cunnington, Rick 

Pacific Scientific Corp Day, Bob 

Pacific Scientific Corp Greenslade, John T. 

Pacific Scientific Corp LaFrance, Bob 

Pacific Scientific Corp Schuman, Alex 

Pacific Scientific Corp Smith, Bob 

Pacific Scientific Corp Spomer, Ed 

Pacific Scientific Corp Todd, Michael C. 

Pacific Scientific Corp VonDerAhe, Ken 

Pacific Scientific Corp Walsh, Tom 

Quantic Industries, Inc Willis, Kenneth E. 

Reynolds Industries Systems, Inc Varosh, Ron 

Rockwell International Corp Cascadden, Raymond 

Rockwell International Corp Erazo, Anibal 

Sandia National Laboratories Andrews, Larry A. 

Sandia National Laboratories . . . Bickes Jr., Robert W. 

Sandia National Laboratories Bonzon, Lloyd L. 

Sandia National Laboratories .... Brigham, William P. 

Sandia National Laboratories Chow, Weng W. 

Sandia National Laboratories Cooper, Paul 

Sandia National Laboratories Curtis, William 

Sandia National Laboratories Fleming, Kevin J. 

Sandia National Laboratories GrubelicH, M. C. 

Sandia National Laboratories Harlan, Jere G. 

Sandia National Laboratories .... Holswade, Scott C. 

Sandia National Laboratories Kass, William 

Sandia National Laboratories Merson, John A. 

Sandia National Laboratories Metzinger, Kurt 

Sandia National Laboratories Mitchell, Dennis E. 

Sandia National Laboratories Salas, F. Jim 

Sandia National Laboratories Setchell, Robert E. 

Sandia National Laboratories .... Stichman, Dr. John 

Sandia National Laboratories Troh, Wayne M. 

Sandia National Laboratories Williams, T. J. 

Santa Barbara Research Center Gonzales, Roman 

Santa Barbara Research Center Jeter, James 

SCB Technologies, Inc McCampbell, C. B. 

Schimmel Company Schimmel, Morry L. 

Special Devices, Inc Sipes, William J. 

Special Devices, Inc Telle, Dennis 

Stresau Company Stresau, Richard 

Teledyne McCormick Selph Ingnam, Robert W. 

Teledyne McCormick Selph Marshall, Clyde 

Teledyne McCormick Selph Smith, Brian 

TPL Inc Brown, Ralph 

United Technologies/USBI Webster, Charles F. 

Universal Propulsion Co., Inc Barlog, Stan 

Universal Propulsion Co., Inc. . . . Magenot, Michael C. 

Universal Propulsion Co., Inc Mayville, Wayne 

Universal Propulsion Co., Inc Wergen, Tom 

University of Notre Dame Gonthier, Keith A. 

University of Notre Dame Powers, Joseph M. 

USAF/30 SW/SESX Gotfraind, Mark 

USAF/45SPW/SESE Wadzinski, Mike 

USAF/B-1B Engineering Branch Tipton, Steve 

USAF/Ogden Air Logistics Center .... Kangas, Charles 
UTC/Chemical Systems Division Lai, K. S. 

ZZY2X Potter, Jed 



Larry A. Andrews 

Sandia National Laboratories 

P.O. Box 5800 

Albuquerque, NM 87185-5800 

J ay a Bajpayee 

NASA Goddard Space Flight Center 

Wallops Flight Facility - Bldg N-159 

Wallops Island, VA 23337 

(804) 824-2374 (FAX) 824-1 51 8 

James Barker 

Halliburton Energy Services 

Explosive Products Center 

2001 S. I35 

Alvarado, TX 76009 

(817)783-5111 (FAX) 783-5812 

Stan Barlog 

Universal Propulsion Co., Inc. 

25401 North Central Ave. 

Phoenix, AZ 85027-7837 

(602) 869-8067 (FAX) 869-8176 

Thomas M. Beckman 

EG&G Mound Applied Technologies 

P.O. Box 3000 

Miamisburg, OH 45343-0987 

(513)865-4551 (FAX) 865-3491 

Larry Bement 

NASA Langley Research Center 

Code 433 

Hampton, VA 23681-0001 

(804) 864-7084 (FAX) 864-7009 

Robert W. Bickes Jr. 

Sandia National Laboratories 

P.O. Box 5800 

Albuquerque, NM 87185-0326 

(505) 844-0423 (FAX) 844-5924 

Tom Blachowski 

Naval Surface Warefare Center 

Indian Head Division - Code 5240E 

101 Strauss Ave. 

Indian Head, MD 20640-5035 

(301) 743-4243 (FAX) 743-4881 



Lloyd L. Bonzon 

Sandia National Laboratories 

P.O. Box 5800 

Albuquerque, NM 87185-0327 

(505) 845-8989 (FAX) 844-5924 

William P. Brigham 

Sandia National Laboratories 

P.O. Box 5800 

Albuquerque, NM 87185-0327 

(505) 845-9107 (FAX) 844-0820 

Ralph Brown 
TPL Inc. 

Raymond Cascadden 

Rockwell International Corporation 

201 N. Douglas St. 

El Segundo. CA 90245 

(310)414-1655 (FAX) 414-2077 

Weng W. Chow 

Sandia National Laboratories 

Division 2235 

P.O. Box 5800 

Albuquerque, NM 87185-5800 

(505)844-9088 (FAX) 844-8168 



David A. Cole 
Hercules Aerospace 
P.O. Box 210 
Rocket Center, WV 
(304) 726-5489 



26726 
(FAX) 726-4730 



Paul Cooper 

Sandia National Laboratories 

M/S 1 1 56 

P.O. Box 5800 

Albuquerque, NM 87185-5800 

(505) 845-7210 (FAX) 845-7602 

Rick Cunnington 

Pacific Scientific Corp. 

Energy Dynamics Division 

7073 West Willis Road, Box 5002 

Chandler, AZ 85226-5111 

(602) 796-1 100 (FAX) 796-0754 



A-2 



William Curtis 

Sandia National Laboratories 

P.O. Box 5800 

Albuquerque, NM 87185 

(505) 845-9649 (FAX) 844-4616 



Werner A. Gans 
Consultant 
1015 Lanark Ct. 
Sunnyvale, CA 94089 
(408) 245-2857 



Bob Day 

Pacific Scientific Corp. 

7073 West Willis Road, Box 5002 

Chandler, AZ 85226-5111 

Robert L. Dow 

Attenuation Technology, Inc. 

9674 Charles Street 

La Plata, MD 20646 

(301) 934-3725 (FAX) 934-3725 

Anibal Erazo 

Rockwell International Corporation 

201 N. Douglas St. 

El Segundo, CA 90245 

(310) 647-2756 (FAX) 647-6824 

Jeff Filliben 

John Hopkins University - CPIA 

10630 Little Patukxent Parkway 

Suite 202 

Columbia, MD 21044-3200 

(410) 992-7305 (FAX) 730-4969 

Kevin J. Fleming 

Sandia National Laboratories 

P.O. Box 5800 

Albuquerque, NM 87185-0327 

(505) 845-8763 (FAX) 844-0820 

Mark Folsom 

Consultant 

25747 Carmel Knolls Drive 

Carmel, CA 93923 

(408) 626-8252 (FAX) 626-1652 

James Gageby 

The Aerospace Corporation 

MS M4/907 

2350 E. El Segondo Blvd. 

El Segondo, CA 90245 

(310) 336-7227 (FAX) 336-1474 



Selma Goldstein 

The Aerospace Corporation 

MS M4-907 

P.O. Box 92957 

Los Angeles, CA 90009-2957 

(310) 336-1013 (FAX) 336-1474 

Keith A. Gonthier 

University of Notre Dame 

Aerospace & Mechanical Engineering 

365 Fitzpatrick Hall 

Notre Dame, IN 46556-5637 

(219) 239-6426 (FAX) 239-8341 

Roman Gonzales 

Santa Barbara Research Center 

Hughes Aircraft Company 

75 Coromar Drive B30/12 

Goleta, CA 93117 

(805) 562-7705 (FAX) 562-7882 

Mark Gotfraind 

U. S. Air Force/30th Space Wing 

30 SW/SESX 

922 N. Brian St. 

Santa Maria, CA 93454 

(805) 928-9637 

John A. Graham 

Ensign Bickford Aerospace Co. 

640 Hopmeadow St. 

P.O. Box 427 

Simsbury, CT 06070 

(203) 843-2325 

John T. Greenslade 

Pacific Scientific Corp. 

Energy Dynamics Division 

7073 West Willis Road, Box 5002 

Chandler, AZ 85226-5111 

(602) 796-1100 (FAX) 796-0754 

M. C. Grubelich 

Sandia National Laboratories 

P.O. Box 5800 

Albuquerque, NM 87185-0326 

(505) 844-9052 (FAX) 844-4709 



A-3 



David Hansen 

Morton International Inc. 

M/SX1830 

3350 Airport Road 

Ogden, UT 84405 

(801) 625-9222 (FAX) 625-4949 

Jeff Hansen 

Aerojet Propulsion Division 

Bldg. 201 9A2, Dept. 5274 

P.O. Box 13222 

Sacramento. CA 95813-6000 

(916) 355-6102 (FAX) 355-6543 

Jere G. Harlan 

Sandia National Laboratories 

P.O. Box 5800 

Albuquerque, NM 87185-0329 

(505) 844-4401 (FAX) 844-4709 

Jeffery L. Hinds 

Naval Surface Warfare Center 

Indian Head Division - Code 520 

101 Strauss Ave. 

Indian Head, MD 20640-5035 

(301)743-6530 (FAX) 743-4881 



James Jeter 

Santa Barbara Research Center 

Hughes Aircraft Company 

75 Coromar Drive 

Goleta, CA 93117 

(805) 562-7539 (FAX) 562-7740 

Charles Kangas 

USAF/Ogden Air Logistics Center 

OO-ALC/LIWCE 

6033 Elm Lane 

Hill AFB, UT 84056 

(801)777-4135 (FAX) 777-9484 

William Kass 

Sandia National Laboratories 

Division 2234 

P.O. Box 5800 

Albuquerque, NM 87185-5800 

(505)844-6844 (FAX) 844-8168 

James E. Kennedy 

Los Alamos National Laboratory 

MS - P950 

Los Alamos, NM 87545 

(505) 667-1468 (FAX)667-6301 



Lane Hinkle 

Martin Marietta Specialty Components 

P.O. Box 2908 

Largo, FL 34649-2908 

(813) 541-8222 (FAX) 545-6757 

William Hoffman 

NASA Johnson Space Center 

Code EP5 

2201 NASA Road One 

Houston, TX 77058 

(713) 483-9056 (FAX) 483-3096 

Scott C. Holswade 

Sandia National Laboratories 

P.O. Box 5800 

Albuquerque, NM 87185-0328 

Robert W. Ingnam 

Teledyne McCormick Selph 

3601 Union Road 

P.O. Box 6 

Hollister, CA 95024-0006 

(408) 637-3731 (FAX) 637-5494 



Daniel Kramer 

EG&G Mound Applied Technologies 

P.O. Box 3000 

Miamisburg, OH 45343-0987 

(513)865-3558 (FAX) 865-3680 

Darrin Krivitsky 

Naval Surface Warfare Center 

Indian Head Division - Code 5240E 

101 Strauss Ave. 

Indian Head, MD 20640-5035 

Bob La France 

Pacific Scientific Corp. 

Energy Systems Division 

7073 West Willis Road, Box 5002 

Chandler, AZ 85226-51 1 1 

(602) 961-0023 (FAX) 961-0577 

K. S. Lai 

UTC/Chemical Systems Division 

P.O. Box 49028 

San Jose, CA 95161-9028 

(408) 776-4327 (FAX) 776-4444 



A-4 



Murphy J. Landry 

Geo-Centers, Inc. 

2201 Buena Vista Dr., SE 

Albuquerque, NM 87106 

(505) 243-3483 (FAX) 242-9497 

Michael C. Magenot 
Universal Propulsion Co., Inc. 
25401 North Central Ave. 
Phoenix, AZ 85027-7837 

Clyde Marshall 

Teledyne McCormick Selph 

8920 Quartz Ave. 

Northridge, CA 91324 

(818) 718-6643 (FAX) 998-3312 

Steve Martin 

Naval Surface Warfare Center 

Indian Head Division 

101 Strauss Ave. 

Indian Head, MD 20640-5000 

(301) 743-4243 (FAX) 743-4881 

Wayne Mayville 

Universal Propulsion Co., Inc. 

25401 North Central Ave. 

Phoenix, AZ 85027-7837 

(602) 869-8067 (FAX) 869-8176 



Pat V. McAllister 
Hercules Inc. 
M/SN1EA1 
P.O. Box 98 
Magna, UT 84044 
(801) 251-6192 



(FAX) 251-6676 



C. B. McCampbell 
SCB Technologies, Inc. 
1009 Bradbury Dr. S.E. 
Albuquerque, NM 87106 

John A. Merson 

Sandia National Laboratories 

P.O. Box 5800 

Albuquerque, NM 87185-0329 

(505) 844-2756 (FAX) 844-4709 



Kurt Metzinger 

Sandia National Laboratories 

Structrual Mechanics Group 1562 

P.O. Box 5800 

Albuquerque, NM 87185-5800 

(505) 844-5077 

Dennis E. Mitchell 

Sandia National Laboratories 

P.O. Box 5800 

Albuquerque, NM 87185-0522 

(505) 845-8656 (FAX) 844-3894 

Joe Moran 

Olin Rocket Research Co. 

P.O. Box 97009 

Redmond, WA 98073-9709 

(206) 885-5000 (FAX) 882-5744 

Jerry Motley 

Halliburton Energy Services 

Explosive Products Center 

2001 S. I35 

Alvarado, TX 76009 

(817) 783-5111 (FAX) 783-5812 

Alan C, Munger 

EG&G Mound Applied Technologies 

P.O. Box 3000 

Miamisburg, OH 45343-0987 

(513) 865-3544 (FAX) 865-3491 

Don Novotny 

Hi-Shear Technology Corp. 

24225 Gamier Street 

Torrance, CA 90509-5323 

(310) 784-7857 (FAX) 325-5354 

Don Nowakowski 

Conax Florida Corporation 

2801 75th Street North 

St. Petersburg, FL 33710 

(813)345-8000 FAX: 345-4217 

Gerald L. O'Barr 
6441 Dennison St. 
San Diego, CA 92122 
(619)453-0071 



A-5 



Peter Ostrowski 

Energetic Materials Technology 

P.O. Box 6931 

Alexandria, VA 22306-0931 

(703) 780-5854 (FAX) 780-4955 

James E. Parenzan 

McDonnell Douglas Space Systems Co. 

MSA3/11-1/L292 

5301 Bolsa Ave. 

Huntington Beach, CA 92647 

(714)896-5778 (FAX) 896-1597 

Jed Potter 

ZZYZX 

25341 Via Oriol 

Valencia, CA 91355 

(805) 259-9491 

Joseph M. Powers 

University of Notre Dame 

Aerospace & Mechanical Engineering 

365 Fitzpatrick Hall 

Notre Dame, IN 46556-5637 

(219)631-5978 (FAX) 631-8341 

Larry Rayburn 

NASA Kennedy Space Center 

NASA Shuttle Pyrotechnic Engineering 

Code TV. MS D-23 

Kennedy Space Center, FL 32899 

(407)861-3652 (FAX) 867-2167 

Steven L. Renfro 

Ensign Bickford Aerospace Co. 

640 Hopmeadow St. 

P.O. Box 427 

Simsbury. CT 06070 

(203) 843-2403 (FAX) 843-2621 

Arthur D. Rhea 

Ensign Bickford Aerospace Co. 

640 Hopmeadow St. 

P.O. Box 427 

Simsbury, CT 06070 

(203) 843-2360 (FAX) 843-2621 

Bill Richardson 

Morton International Inc. 

M/SX1870 

3350 Airport Road 

Ogden, UT 84405 

(801) 625-8222 (FAX) 625-4949 



Steven P. Robinson 

Boeing Defense & Space Group 

M/S 81-05 

P.O. Box 3999 

Seattle, WA 98124-2499 

(206) 773-1894 (FAX) 773-4846 

William W. St. Cyr 

NASA Stennis Space Center 

Code KA22 Building 1100 

Stennis Space Center, MS 39529-6000 

(601 ) 688-1 1 34 (FAX) 688-33 1 2 

F. Jim Salas 

Sandia National Laboratories 

P.O. Box 5800 

Albuquerque, NM 87185-0329 

(505) 844-3265 (FAX) 844-4709 

Morry L. Schimmel 
Schimmel Company 
8127 Amherst Avenue 
St. Louis, MO 63130 
(314)863-7725 (FAX) 727-8107 

Jan N. Schlamp 

Naval Surface Warfare Center 

Code 4073 

300 Highway 361 

Crane, IN 47522-5001 

(812) 854-5431 (FAX) 854-171 1 

Norman R. Schulze 
National Aeronautics 

and Space Administration 
Code QW 

Washington, DC 20546 
(202) 358-0537 (FAX) 358-2778 

Alex Schuman 

Pacific Scientific Corp. 

Energy Dynamics Division 

Box 750 

Litchfield Park. AZ 85340 

(602) 932-8409 (FAX) 932-8949 

Tom Seeholzer 

NASA Lewis Research Center 

Code 4330 M/S 86-10 

21000 Brookpark Road 

Cleveland, Ohio 44135 

(216) 433-2523 FAX 433-6382 



A-6 



Robert E. Setchell 

Sandia National Laboratories 

Department 5166 

P.O. Box 5800 

Albuquerque, NM 87185-0445 

(505) 844-3847 (FAX) 844-7431 

William J. Sipes 

Special Devices, Inc. 

16830 West Placerita Canyon Road 

Newhall, CA 91321 

(805) 259-0753 (FAX) 254-4721 



Dr. John Stichman 
Sandia National Laboratories 
Surety Components and 
Instrumentation Center 
P.O. Box 5800 
Albuquerque, NM 87185-5800 

Richard Stresau 
Stresau Company 
Star Route 
Spooner, Wl 54801 
(715)635-8497 



Bob Smith 

Pacific Scientific Corp. 

102 South Litchfield Road 

Goodyear, AZ 85338-1295 

(602) 932-8450 (FAX) 932-8949 



Dennis Talle 

Special Devices, Inc. 

16830 West Placerita Canyon Road 

Newhall, CA 91321 

(805) 259-0753 (FAX) 254-4721 



Brian Smith 

Teledyne McCormick Selph 

3601 Union Road 

P.O. Box 6 

Hollister, CA 95024-0006 

(408) 637-3731 (FAX) 637-5494 



Ramie Thompson 

Franklin Applied Physics 

98 Highland Ave. 

P.O. Box 313 

Oaks, PA 19456 

(215) 666-6645 (FAX) 666-0173 



Floyd Smith 

Analex Corporation 

3001 Aerospace Parkway 

Brookpark, OH 44142-1003 

(216) 977-0201 (FAX) 977-0200 



Michael Tierney 

McDonnell Douglas 

10171 Halawa Dr. 

Huntington Beach, CA 92646 

(714) 963-7242 (FAX) 896-6995 



Ed Spangler 

EG&G Mound Applied Technologies 

P.O. Box 3000 

Miamisburg, OH 45343-3000 

(513) 865-3528 (FAX) 865-3491 

Ed Spomer 

Pacific Scientific Corp. 

Energy Dynamics Division 

102 South Litchfield Road 

Goodyear, AZ 85338-1295 

(602) 932-8100 (FAX) 932-8949 

Paul Steffes 

Analex Corporation 

3001 Aerospace Parkway 

Brookpark, OH 44142-1003 

(216) 977-0123 (FAX) 977-0200 



Steve Tipton 

USAF/B-1B Engineering Branch 

Air Logistics Center (AFMC) 

3001 Staff Drive, STE 2AA86A 

Tinker AFB, OK 73145-3006 

(405) 736-7444 (FAX) 736-3714 

Michael C. Todd 

Pacific Scientific Corp. 

Energy Systems Division 

7073 West Willis Road, Box 5002 

Chandler, AZ 85226-5111 

(602) 796-1 100 (FAX) 796-0754 

Wayne M. Troh 

Sandia National Laboratories 

M/S 1512 

P.O. Box 5800 

Albuquerque, NM 87185-5800 

(505) 844-9516 (FAX) 844-8251 



A-7 



Ron Varosh 

Reynolds Industries Systems, Inc. 

3420 Fostoria Way 

San Ramon, CA 94583 

(510) 866-0650 (FAX) 866-0564 

Ken VonDerAhe 

Pacific Scientific Corp. 

Energy Systems Division 

7073 West Willis Road, Box 5002 

Chandler, AZ 85226-5111 

(602) 796-1 100 (FAX) 796-0754 

Mike Wadzinski 

USAF/45SPW/SESE 

1201 Minuteman St. 

Patrick Air Force Base, FL 32925 

(407) 494-7629 (FAX) 494-6535 

Tom Walsh 

Pacific Scientific Corp. 

7073 West Willis Road, Box 5002 

Chandler, AZ 85226-5111 

Bruce Watson 

Olin Rocket Research Co. 

P.O. Box 97009 

Redmond, WA 98073-9709 

(206) 885-5000 (FAX) 882-5804 

Charles F. Webster 

United Technologies/USBI 

M/S EN 

P.O. Box 1900 

Huntsville, AL 35811 

(205) 721-2342 (FAX) 721-2263 

Richard Webster 

Hi-Shear Technology Corp. 

24225 Gamier Street 

Torrance, CA 90509-5323 

(310) 784-7867 (FAX) 325-5354 

Tom Wergen 

Universal Propulsion Co., Inc. 

25401 North Central Ave. 

Phoenix, AZ 85027-7837 

(602) 869-8067 (FAX) 869-8176 



Ian Whalley 

McDonnell Douglas Aerospace 

M/S A3-L292/11-1 

5301 Bolsa Ave. 

Huntington Beach, CA 92647 

(714) 896-6491 (FAX) 896-1 106 

T. J. Williams 

Sandia National Laboratories 

P.O. Box 5800 

Albuquerque, NM 87185-5800 

(505) 844-3356 (FAX) 844-4616 

Kenneth E. Willis 

Quantic Industries, Inc. 

900 Commercial Street 

San Carlos, CA 94070-4084 

(415) 637-3074 (FAX) 592-4669 

Steven T. Wirrig 

Inland Fisher Guide Division of GM 

250 Northwoods Blvd. M/S 110 

Vandalia, OH 45377 

(513) 356-2271 (FAX) 356-2280 

T. Eric Wong 

The Aerospace Corporation 

MS M4/901 

2350 E. El Segondo Blvd. 

El Segondo, CA 90245 

(310) 336-6190 (FAX) 336-1474 

Lance Wood 

Martin Marietta Astronautics 

Mail Stop 5450 

P.O. Box 179 

Denver, CO 80201 

(303) 971-1218 (FAX) 977-1940 

Ron Ziegler 

American Safety Flight Systems, Inc. 
11605 Rivera Road, NE 
Albuquerque, NM 87111-5336 
(505) 294-1645 (FAX) 294-1645 



A-8 



REPORT DOCUMENTATION PAGE 



Form Approved 
OMB No. 0704-01 8S 



PiAKc ; reporting burden for this cofection of information is estimated to ewerage 1 hour per response, including tfw lime tor reviewing instructions, searching existing daU sources, 
gathering and maintaining the data needed, and completing and reviewing the cottedion of information. Send comments regarding this burden estimate or any other aspect of this 
collection of Information, including suggestions for reducing this burden, to Washington Headquarters Services, Directorate lor Information Operations and Reports. 1215 Jefferson 
Davis Highway. Suite 1204. Arlington. VA 22202-4302. and to the Office of Management and Budget Paperwork Redtiction Pro^ (07^ 20503. 



1. AGENCY USE ONLY (Leave blank) 



4. TITLE AND SUBTITLE 



2. REPORT DATE 

February L994 



3. REPORT TYPE AND DATES COVERED 

CP February 8-9, 1995 



Second NASA Aerospace Pyrotechnics Systems Workshop 



6. AUTHOR(S) 

William W. St. Cyr, compiler 



7. PERFORMING ORGANIZATION NAME(S) AND ADDRESSES) 

National Aeronautics and Space Administration 

John C. Stennis Space Center 

CodeKA60 Building 1100 

Stennis Space Center, MS 39529-6000 



5. FUNDING NUMBERS 



8. PERFORMING ORGANIZATION 
REPORT NUMBER 



CP 3258 



9. SPONSORING/MONITORING AGENCY NAME(S) AND ADDRESSES) 

Office of Safety and Mission Quality 

CodeQW 

NASA Headquarters 

Washington, D.C. 20546 



10. SPONSORING/MONITORING 
AGENCY REPORT NUMBER 



11. SUPPLEMENTARY NOTES 



Hosted by Sandia National Laboratories 
Albuquerque, New Mexico 



12a. DISTRIBUTION/AVAILABILITY STATEMENT 

Unlimited 



12b. DISTRIBUTION CODE 



13. ABSTRACT (Maximum 200 words) 

This NASA Conference Publication contains the proceedings of the Second NASA Aerospace Pyrotechnics Systems 

Workshop held at Sandia National Laboratories, Albuquerque, New Mexico, February 8-9, 1994. The papers are grouped 

by sessions: 

Session 1 - Laser Initiation and Laser Systems 

Session 2 - Electric Initiation 

Session 3 - Mechanisms & Explosively Actuated Devices 

Session 4 - Analytical Methods and Studies 

Session 5 - Miscellaneous 

A sixth session, a panel discussion and open forum, concluded the workshop. 



14. SUBJECT TERMS 



Pyrotechnics, laser ordnance, safety, pyrotechnic database 

and catalogue, electric initiation, explosively actuated devices, laser 

safe and arm system, mechanisms, modeling of pvros. 



15. NUMBER OF PA^ES 

3M. 



16. PRICE CODE 



17. SECURITY CLASSIFICATION 
OF REPORT 

Unclassified 



18. SECURITY CLASSIFICATION 
OF THIS PAGE 

Unclassified 



19. SECURITY CLASSIFICATION 
OF ABSTRACT 

Unclassified 



20. LIMITATION OF ABSTRACT 



NSN 7540-01-280-5500 



Standard Form 298 (Rev. 2-89)